Fannie and Freddie turned out to be almost as clueless as your average investor. They, too, relied on the rating agencies, although Fannie did so with a tiny bit of caution. (“Although we invest almost exclusively in triple-A-rated securities, there is a concern that rating agencies may not be properly assessing the risk in these securities,” noted a Fannie internal document in the spring of 2005.) Not enough caution, however. After the crisis, HUD would report that the value of the Wall Street-created securities owned by Fannie and Freddie fell as much as 90 percent from the time of purchase.
The GSEs also began buying, guaranteeing, and selling those not-quite-subprime Alt-A mortgages. Fannie executives insist that they never bought or guaranteed more than a few billion dollars worth of loans they considered subprime. They never guaranteed loans with layered risks, for instance. But many of the borderline loans they guaranteed would certainly be categorized as subprime by others in the marketplace. To this day, former Fannie Mae executives will insist that they chose the securities they guaranteed more carefully than others. And maybe they did; after the crisis, Fannie and Freddie defenders would point out that in every mortgage category, from prime to Alt-A to subprime, the GSEs’ loans defaulted at rates below the national average.
But just as with the purchase of triple-A securities, guaranteeing Alt-A loans had little to do with housing goals and everything to do with profits and market share. They were simply more profitable than guaranteeing thirty-year fixed loans. “We were lured into it by the big margins,” says a former executive. Both companies got warnings about the true state of market—Fannie from the outside and Freddie from the inside. Michelle Leigh, a vice president at IndyMac, later claimed in a lawsuit that she tried to warn Fannie about the Alt-A loans it was buying from IndyMac, which were riddled with problems. Fannie didn’t respond, and increased its purchases of loans from the company, according to the lawsuit.
Over at Freddie, chief credit officer David Andrukonis warned the company’s new CEO, Dick Syron, the former chairman of the Boston Federal Reserve, about the riskiness of no-income, no-asset loans. (They were called NINA loans.) “Freddie Mac should withdraw from the NINA market as soon as practical,” Andrukonis wrote in the fall of 2004. “Today’s NINA appears to target borrowers who would have trouble qualifying for a mortgage if their financial position were adequately disclosed.” He added, “What better way to highlight our sense of mission than to walk away from profitable business because it hurts the borrowers we are trying to serve?”
Between 2005 and 2007, about one in five mortgages Fannie and Freddie purchased or insured was Alt-A or subprime, according to a study by Jason Thomas. By the end of 2007, Fannie Mae had $350 billion in Alt-A exposure and another $166 billion in exposure to mortgages that it defined as subprime or whose recipients had FICO scores of less than 620. Freddie had $205 billion in Alt-A exposure and $173 billion in exposure to subprime or sub-620 FICO scores. Thomas calculates that that meant the GSEs owned about 23 percent of the subprime mortgage-backed securities outstanding at that time and a whopping 58 percent of the total Alt-A mortgages outstanding.
There was no worse time to accumulate exposure to Alt-A and subprime loans than the 2005 to 2007 time period. Some critics would later point to these massive purchases in an effort to blame the entire crisis on Fannie and Freddie. But as Thomas points out, it’s precisely because they were so late to the party that their losses would be so immense.
Another irony is that, in the end, OFHEO, despite its brief stance as an aggressive regulator, failed as miserably as the GSEs. As Raines would later point out, “Fannie and Freddie succumbed to the pressure, and they did so right in front of OFHEO.” After the accounting scandal, OFHEO had examiners in Fannie’s offices on a full-time basis. There was very little that Fannie Mae did that OFHEO didn’t know about. OFHEO’s 2006 report to Congress had a cover letter that read in part, “OFHEO is working with the Enterprises to provide guidance on subprime … mortgages.” OFHEO had the right to suspend the affordable housing goals if the agency felt they threatened the GSEs’ capital position. At any moment along the way, OFHEO could have stopped the GSEs from buying risky loans by citing “safety and soundness” concerns. But it didn’t. Like the other regulators who were charged with looking after the health of the financial system, OFHEO simply didn’t appreciate the credit risk until it was too late.
The last, and most painful, irony is that the two longtime rival armies in the securitization market—the investment banks and the GSEs—would end up magnifying each other’s sins rather than keeping each other in check. Without the GSEs’ buying power, the private market would never have been as big as it got. And without Wall Street, there never would have been all those bad mortgages for the GSEs to binge on.
Which is why some Fannie defenders argue that the GSEs, rather than being the villains of the crisis, were really the victims. That may be, but they were far from innocent victims.
13
The Wrap
His temper. That’s what AIG-FP traders always mentioned whenever they talked about their old boss, Joe Cassano. Yes, they would also take note of his wide-ranging intelligence, and the way he knew every FP employee’s name, and the pleasure he took in handing out multimillion-dollar bonuses each December—“like Santa Claus,” recalls a former executive. But his temper tended to dominate conversations about him, because that’s what everyone at FP had to cope with every day.
It was brutal and indiscriminate—“terrifying when unleashed,” says an ex-trader. “Sometimes he could seem uncontrollable.” Cassano would rage at traders who were making the company a fortune and traders who were on a losing streak. He would go out of his way to embarrass executives in front of their peers, and blow up over the most inconsequential things. “Talking to him was like walking on eggshells,” says another former FP executive. “You were always worried about what would set him off.”
Once, he got mad because a trader wore a V-neck sweater over a T-shirt. From two desks over, he loudly berated the man and then sent him home to change into a collared shirt. A new hire, speaking to Cassano for the first time, told him that the firm was making 50 basis points on a certain $1 billion transaction. “I said that was $5 million a year,” recalls the trader. Cassano erupted: “How dare you do the math on me!” He was a bully. It was his fatal flaw.
Cassano had taken over AIG-FP when Tom Savage retired in 2001. Though he had been Savage’s top deputy, it was no sure thing that Cassano would get the top job. But Hank Greenberg had taken a shine to Cassano; a tough, up-from-the-streets manager who cared about AIG to the exclusion of all else, Cassano surely reminded Greenberg of himself. It also helped that Cassano never forgot who was boss. “Joe managed Hank beautifully,” says a former colleague.
Cassano had learned the derivatives trade from the ground up, having begun his FP career in the back office. In the late 1990s, when the London office was floundering, Savage moved him there to fix the situation. He did, making London his home and traveling to the Connecticut office one week a month. When AIG-FP first began selling credit default swaps, Cassano ran the business. He may have lacked a degree in high finance, but nobody could say he didn’t know derivatives.
He also cared deeply about the business. His former employees all stress that as well. Whenever an executive said a deal couldn’t be done because of some deficiency with FP, Cassano would respond with fury. “He took that kind of thing personally,” says a former trader. “In his eyes, you were blaming his company for why you were ineffective.” In his mind, FP had no deficiencies. Cassano’s devotion to FP was also why his former colleagues all say that he would never, ever do anything that he thought might damage it. Which, in retrospect, was the most surprising thing about him. Joe Cassano was positively risk averse.
Ever since Howard Sosin’s departure in 1993, Greenberg had insisted that FP executives defer half their compensation to protect against deals later going sour. Cassano himself went much further. In 2007, for
instance, he was paid $38 million, but pulled out only $1.25 million, keeping the rest in his deferred compensation pool. He had no incentive to take foolish risks. Whenever FP devised a new product, he took it to Greenberg to get the CEO’s blessing. He made traders pull back from positions he thought were becoming too risky. “The company took minimal risk,” says one former trader. When deals were brought to him, Cassano would pick them apart, looking for hidden risks. “He would say, ‘Be careful out there. Don’t take big positions,’” recalls the former trader. “He wasn’t a cowboy.”
“It was an extremely well-run business,” this trader continues. “But there was one blind spot.”
The blind spot was AIG-FP’s credit default swap business. Ever since that original BISTRO deal with J.P. Morgan, FP had created a profitable niche by taking on the risk of insuring the super-senior tranches of CDOs. These were the top-tier, triple-A tranches, the ones that got hit only if all the lower tranches got wiped out. Although it had been a sexy little business in the beginning, with innovative deals and healthy margins, it had become fairly humdrum. For the first four or five years, the credit default swap business was focused primarily on corporate credits—first writing protection against the possibility that a particular company’s debt might default, and then insuring the super-senior tranches of CDOs that were primarily made up of corporate loans. But as competitors entered the business, the spreads narrowed and the profits dwindled. Soon, FP’s traders were looking for new ways to use credit derivatives to make money.
FP found several things. It created a business aimed at helping banks—primarily European banks—evade their capital rules. AIG-FP sold credit default swaps to banks that held a variety of highly rated paper; the FP wrap, as it was called, allowed them to decrease their capital. There was nothing illegal about this; seeking “capital relief” had become widespread ever since the adoption of risk-based capital standards. Nor did FP hide what it was doing. It called the business—appropriately enough—“regulatory capital.” By the time of the financial crisis, AIG-FP had insured around $200 billion of highly rated bank assets.
And in 2004, AIG-FP began selling credit protection on triple-A tranches of a new kind of CDO, called a multisector CDO. In this, AIG-FP was following the evolution of the CDO business itself, which had gone from BISTRO—a CDO made up of one bank’s corporate loan portfolio—to CDOs that consisted of disparate corporate credits, to this new multisector CDO. Multisector CDOs were “highly diversified kitchen sinks,” as one FP trader put it, that included everything from student loans to credit card debt to prime commercial real estate mortgage-backed securities to a smattering of subprime residential mortgage-backed securities. The theory, as always, was that diversification would protect against losses; the different asset classes in a multisector CDO were supposed to be uncorrelated. A Yale economist named Gary Gorton was hired to work up the risk models, which showed—naturally!—that the possibility of losses reaching the super-senior tranches was so tiny as to be nearly nonexistent. To FP’s executives, wrapping the super-seniors felt like free money.
The executive who marketed credit default swaps for AIG-FP was Al Frost. Many FP executives were wary of him, viewing him as someone who had a way of shoehorning his way into businesses started by others. But he was a Cassano favorite and a member of the boss’s inner circle. Like everyone else, he feared Cassano’s temper. “He was more worried about Joe’s temper than about bringing him straight information,” says someone who worked with him.
Frost was a glad-hander, with friends up and down Wall Street. As these new multisector CDOs were being developed, Frost was among the people at FP who soon realized there was a need for someone to wrap the triple-A tranches—a need that AIG-FP was uniquely capable of filling. AIG-FP by then was an experienced, trusted credit default swap counterparty that knew the ins and outs of swap contracts. And FP’s swaps were especially appealing to underwriters because they were backed by the parent company’s own stellar triple-A rating.
Or were they? Here was a question that no one at FP—or its counterparties—ever thought much about: What did it really mean to be backed by AIG’s triple-A rating? It most certainly did not mean that the parent company’s capital reserves were at FP’s disposal. AIG was an insurance company; there were severe limits as to how it could deploy its capital reserves. Says a former AIG executive: “We had capital, but it wasn’t mobile.” Much of AIG’s capital was walled off in the company’s insurance units, where it could only be used to shore up that particular division. Surprisingly little of it could be moved to FP, even in a crisis. In truth, despite being owned by the world’s largest insurance company, AIG-FP was really a stand-alone derivatives dealer. The parent company did not have its back.
“If you run a derivatives company,” says a former AIG-FP executive, “all you have to do is be wrong once, given the amount of leverage and the size of the book.” Hank Greenberg viewed AIG’s triple-A rating as critical to his business model, yet he never realized that the very existence of AIG-FP put that triple-A rating at risk.
That same uncritical belief in the strength of AIG’s triple-A was the reason FP never bothered to hedge its exposure to the super-senior tranches it was insuring. This was a sharp departure from the company’s usual practice. But AIG executives felt that because the deals were deemed to be riskless, “why would you hedge a riskless transaction? And if you did hedge,” says a former FP trader, “you would be getting protection from someone who was unlikely to be around, because it was a lesser entity.” In other words, any financial event powerful enough to cause losses to the super-seniors was also likely to bring down the counterparties.
Written into FP’s contracts were so-called collateral triggers, which allowed counterparties to demand that AIG put up collateral—that is, cold, hard cash—if certain events took place. One trigger was a drop in AIG’s credit rating to single-A. A second trigger was a downgrade of the super-seniors. And the third trigger was a decline in the market value of the securities AIG had wrapped—even if those securities retained their triple-A rating.
It is hard to know for sure if these triggers were there from the start. Frost ran his department like a little fiefdom; he tended to impart information on a need-to-know basis. (Through his attorney, Frost denies that he didn’t talk freely about what was going on in his business.) AIG-FP’s chief competitors in wrapping triple-A CDO tranches were the so-called monoline insurers, like MBIA and Ambac. Their business model did not allow for collateral triggers. AIG-FP’s willingness to agree to the triggers gave it a big marketing advantage.
It is also unclear who else in the company knew about the triggers. Cassano knew, of course. Greenberg says he knew as well, but they didn’t trouble him: “At the time, the business was so small, and besides, we had the triple-A.” But almost nobody else at AIG appears to have known about them—not AIG’s risk managers, not the executives who oversaw Cassano and FP, and not other FP executives. Every quarter, AIG-FP sent a memorandum to AIG management, updating its positions and exposure. The collateral calls were never a part of those memos, according to former AIG executives. “I don’t think Joe ever really focused on them. It was just another facet of the deal,” says one executive. Cassano and Frost appear to have assumed it was simply not possible that they would ever have to put up collateral. In the days when Sosin and then Savage had run AIG-FP, it is likely that the collateral triggers would have been discussed openly with FP and the AIG executive suite. But Cassano’s FP was a much more secretive place; former AIG executives say they had to practically conduct interrogations to pry even the most mundane information out of FP. It wasn’t until 2007, when Cassano acknowledged the collateral triggers on a conference call with investors, that most AIG executives first found out about them.
“It was startling news,” says a former trader. “It was the stupidest thing we ever did.”
By the time FP was getting into the multisector CDO business, Hank Greenberg was in his late seventi
es. He still didn’t have a succession plan. He still ruled AIG with an iron fist. And he was still heralded as the Great Man of modern insurance. Even though by 2004 AIG ranked tenth on the Fortune 500, with almost $100 billion in revenue and more than $11 billion in profits, Greenberg still managed to churn out 15 percent earnings gains each year. It seemed a miracle.
As ever, the intricacies of AIG remained largely in Hank Greenberg’s head. And those intricacies had become truly bewildering, beyond the ken of most mere mortals. When AIG set up a new subsidiary, it often launched a dozen or more surrounding subsidiaries to take advantage of tax laws, reinsurance possibilities, a whole gamut of small advantages. It had, literally, hundreds of such subsidiaries. “There was a kind of scheming mentality,” says someone familiar with the AIG culture. “They always seemed to be thinking, ‘How do we beat the system?’”
Most CEOs have maybe a half dozen top executives reporting directly to them; Greenberg had nearly thirty direct reports. That meant that only Greenberg and maybe his longtime sidekick Ed Matthews had a complete grasp of AIG’s convoluted businesses. It also meant that nobody besides Greenberg and Matthews understood all the risks the company was taking. “In other firms, there are checks and balances,” says a former AIG consultant.
“So far as I could tell, AIG had no formal risk function.” In effect, Hank Greenberg was AIG’s one-man risk department.
“Hank ran the company unlike any other twentieth-century company,” says another AIG consultant. “Even though it had gotten huge, there was no big company infrastructure. The systems were completely antiquated. It still gathered its earnings data every quarter by hand. And all decisions were made by him to a remarkable degree.”
All the Devils Are Here Page 26
