All the Devils Are Here

by Bethany McLean; Joe Nocera

  To put it another way, while outsiders may have thought they were seeing an earnings machine, insiders saw something that more resembled a Rube Goldberg contraption. Why did AIG have such an antiquated infrastructure? Because Greenberg didn’t like to pay for anything that didn’t generate revenue. So he scrimped on computers and software that other companies bought as a matter of course, and refused to pay the salaries necessary to hire a first-rate operations staff.

  Why did he have thirty direct reports? Because that way he could keep all his division managers segregated from each other. “Hank wanted people to operate with the idea that they had their own little private P&L and they were only responsible for what happened in their cell,” says a former AIG executive.

  Many of AIG’s subsidiaries were also interlocking. So for instance, National Union Fire Insurance Company, a big insurance subsidiary, had a significant ownership stake in ILFC, AIG’s airplane leasing company. An operation called AIG Foreign Life—an overseas life insurance company that was one of the company’s crown jewels—had a large portion of its capital in AIG stock. And of course SICO, the company that Greenberg used to accumulate his top executives’ retirement money and which he himself ran, was AIG’s largest stockholder.

  Although Greenberg obviously was in command of everything that went on at AIG, its formal internal controls and processes—the kind that every public company has in place—were decidedly subpar. AIG’s longtime banker was Goldman Sachs; the firm was constantly involved in AIG deals, and Greenberg had close ties with the firm’s top investment bankers. Yet one former Goldman banker recalls that whenever a new deal was in the works, the junior bankers working on it would find, as he put it, “deficiencies” in AIG’s internal controls. “In the due diligence calls, the company always did end runs around our questions and said they were aware of internal control problems and were working on fixing them. If we weren’t satisfied with the answers, we would have to go to senior relationship people at Goldman and get the go-ahead for the deal. Half the time,” he adds, “the senior people didn’t want to hear any shit about AIG’s problems.” The deals always got done.

  And then there was AIG’s risk taking. Though he had built a very large company, Greenberg still wanted his executives to be risk-taking entrepreneurs in the way they approached their businesses. “He repeatedly made the point that AIG’s balance sheet was so big and so highly rated that AIG could take risks that no other company had the strength to take,” says a former executive. That was the whole point of having a triple-A, after all: AIG could take more risk, and be rewarded for it. “Being that big, and that prone to risk taking—it’s just not a good combination,” says a former AIG executive. “Those two things together do not equal prudence.”

  And it wasn’t just one or two subsidiaries, either. The airline leasing subsidiary, the investment arm, the insurance companies—they were all seeking out risks they could take from leveraging the triple-A rating. Among other consequences, this led a surprising number of AIG divisions to invest in subprime mortgages. FP, of course, was insuring super-senior CDO tranches. But AIG also had a mortgage originator making subprime loans. It had a mortgage insurance unit that was guaranteeing subprime loans. And it had a securities lending program that was investing in subprime mortgages.

  It’s worth pausing for a moment on that last one, to get a sense of just how willing AIG was to push into risky areas. The purpose of any securities lending program is to loan stock from a company’s investment portfolio to short sellers. (A short seller has to borrow stock in order to short it.) The borrower puts up cash as collateral, which the lender is supposed to invest in short-term liquid securities that can be quickly cashed in when the borrower returns the stock. For that reason, most companies with securities lending programs invest the cash in low-yielding, short-term commercial paper. But AIG decided to invest a portion of the money in largely illiquid CDO tranches. Its executives reasoned that since the short sellers would never all ask for their money back at the same time—would they?—they could keep some of the cash in CDOs and turn the wider spreads into bigger profits. The possibility that something might happen someday that would cause all the borrowers to demand their money at the same time was, once again, viewed as implausible. By 2007, the company had $78 billion from the securities lending program tied up in mortgage-backed securities. “It was incredibly irresponsible,” says someone who was there.

  When you got right down to it, there was something almost Wizard of Oz-like about the way AIG was run. It was simply not the company it was purported to be. Behind its impressive facade was a tough, stubborn old man who refused to groom a successor, refused to run his company the way modern companies were supposed to be run, and refused to play by anybody’s rules but his own. Those attributes were about to hurt both him and his beloved AIG.

  AIG’s stock peaked toward the end of 2000, at around $103 a share. It was, by then, a very rich stock, with a much higher valuation than its competitors; keeping it that way was deeply important to Greenberg. But over the next year or so, as the Internet bubble burst, the Enron and WorldCom scandals broke out into the open, and the 9/11 attacks took place, all insurance company stocks suffered, including AIG’s. By 2002, the stock was struggling to stay above $65 a share.

  In retrospect, it appears that these were not the only factors causing the stock to drop. The market was beginning to realize that AIG was going to have increasing difficulty meeting Greenberg’s 15 percent earnings target. Partly, that was simply a function of AIG’s mammoth size: the bigger AIG got, the harder it became to tack on 15 percent in additional profits each year. Analysts call this the law of large numbers.

  There had long been suspicion that Greenberg got to his 15 percent target not just by pushing his executives to take more risk, but by taking advantage of complex accounting rules to help smooth out earnings, disguise underwriting losses, and tuck away surpluses that he could use for a rainy day. AIG had set up a series of reinsurance companies in places like Bermuda and Barbados—away from the prying eyes of U.S. regulators—which, although ostensibly independent, did almost all their business with AIG. Reinsurance companies exist to take on risk that insurers like AIG want to lay off, but state regulators had long suspected that AIG did at least some of its reinsurance deals to pretty up its books. AIG was by no means an Enron—its businesses were very real and so, for the most part, were its profits. But wasn’t it at least a little implausible that, year after year, its earnings never did anything but go up by 15 percent?

  Once, a Wall Street analyst took some clients to see Greenberg and asked him point-blank how he managed to produce that steady stream of earnings growth “in this highly volatile industry,” as he put it. “Aren’t you concerned that the SEC or someone is going to look at AIG and ask if you are managing earnings?”

  Greenberg was not happy with the question, but he gave a surprisingly straightforward answer. “Look,” he replied. “We are in the long-tail liability business”—meaning that, though the risks AIG insured didn’t occur very often, the payout was very large when they did. “If there is one thing we have learned, it is that there are risks we can’t anticipate, so when we have extra capital we are justified in setting it aside.” He added, “What do you want? Do you want steady growth? Or do you want up 60 percent one quarter and down 15 percent the next?” The analyst recalls thinking that Greenberg had just admitted he was managing earnings.

  It was around this same time—with the stock sliding and the market wondering if AIG was finally bumping up against the law of big numbers—that the company found itself in a series of small scandals. The first came in 2001. The SEC accused AIG’s National Union unit of constructing what amounted to a sham insurance transaction to help a company called Brightpoint hide nearly $12 million in losses a few years earlier. “It was some dinkyassed insurance deal that they did that fudged around with the accounting at the end of the day,” says a former executive. And yet it took two years to settle the cas
e—years in which AIG, and Greenberg, infuriated the SEC by dragging its feet on producing documents. “He was arrogant to those people,” recalls this same former executive. “Uncooperative.” When the case was finally settled—with AIG paying a $10 million fine—the SEC went out of its way to criticize the company’s behavior. “The penalty,” the agency said in its press release, “reflects AIG’s participation in the Brightpoint fraud, as well as misconduct by AIG during the Commission’s investigation of this matter.”

  Even as the SEC was prosecuting the Brightpoint deal, the agency opened up a second AIG investigation. This was a much bigger deal, and the guilty party this time was AIG-FP. The SEC eventually charged FP with conducting a series of fraudulent transactions that helped PNC, the big Pittsburgh bank, hide more than $750 million in dubious loans. The allegations were serious enough that the Justice Department opened a criminal investigation. Once again, it took several years to settle the case, this time with AIG agreeing to pay a fine of $80 million, with an additional $46 million in restitution. FP also signed a deferred prosecution agreement with the Justice Department—an extremely serious consequence.

  The fine, however, would have been $20 million less but for the fact that at the last minute Greenberg reneged on the original settlement with the agency, vowing to fight the charges. AIG even issued a press release calling the government’s actions “unwarranted.” The SEC was furious, and the AIG board was shocked. It demanded that Greenberg settle. In addition to the extra $20 million and the deferred prosecution agreement, the government installed a full-time monitor inside AIG—in effect, a government compliance officer.

  In late 2003, with the PNC investigation in full swirl, Cassano met with Greenberg, as he did every year, to show him how he planned to distribute the FP bonuses. As they were going through the numbers, Greenberg said, “Joe, if we’re paying a fine on PNC, it is going to come out of your pocket”—meaning the FP bonus pool. Cassano was startled. “We’re partners,” he replied. “We split things down the middle, good or bad. That’s our agreement.” Greenberg said, “I don’t go for people breaking the law. I’m telling you I’m not picking it up.” Furious now, Cassano flashed Greenberg the temper that the FP traders knew so well. “Go fuck yourself,” he said, pushing his finger in Greenberg’s face. “Calm down, Joe,” said Greenberg. “Don’t get upset.” Cassano did calm down—and Greenberg took the SEC fine out of FP’s bonus pool. There wasn’t a thing Cassano could do about it.

  You would think that, by now, Greenberg would have learned at least one lesson—that a company accused of wrongdoing shouldn’t give the government the back of its hand. All that did was cause the government to dig in its heels. Thanks to Enron in particular, the rules had changed. The kind of “earnings management” that had routinely gone on in corporate America was no longer okay. State insurance regulators, who had long looked the other way at some of AIG’s accounting, were stiffening their spines. But Greenberg, stubborn to the end, expected the same kind of deference from government officials that he routinely received from the corporate world and from his employees. Which is why it was all the more unfortunate that the next government official to go after AIG was Eliot Spitzer, the attorney general of New York. Spitzer, who was preparing to run for governor, reveled in taking on tough guys like Greenberg.

  By 2004, Spitzer had become the scourge of Wall Street. He had exposed conflicts of interest by Wall Street analysts and dug into sleazy behavior in the mutual fund industry. He had sued another tough guy, Richard Grasso, the former head of the New York Stock Exchange, attempting to claw back some of Grasso’s $140 million pay package. Most recently, he had turned his attention to the insurance industry. His original target was Marsh & McLennan, the world’s largest insurance broker, which, it so happened, was run by Greenberg’s eldest son, Jeffrey. In October 2004, brandishing a raft of incriminating e-mails, Spitzer accused the company of conducting a long-running bid rigging scheme. He also said he wouldn’t even begin to talk about a settlement until Jeffrey Greenberg resigned. The Marsh board quickly fired Greenberg. Four months after that, Marsh & McLennan agreed to pay $850 million restitution and apologized for its “shameful” conduct. Among the insurance companies accused of paying kickbacks to Marsh was AIG. Several AIG executives pleaded guilty for their role in the scheme.

  Greenberg remained unrepentant. On February 9, 2005, just days after the Marsh & McLennan settlement was announced, AIG reported its $11 billion 2004 profits. During the earnings call with investors, Greenberg was asked about “the relatively hostile regulatory environment.” He replied, “When you begin to look at foot faults and make them into a murder charge, then you have gone too far.”

  It would be hard to imagine a more poorly timed remark. On February 10, Spitzer got wind of a big reinsurance deal AIG had done with General Re, which was owned by Berkshire Hathaway. The information had come from Gen Re’s lawyers, who had uncovered the deal in the course of another investigation. The purpose of the transaction, Spitzer was told, was to boost AIG’s reserves by $500 million. Wall Street analysts had been calling for AIG to increase its reserve, and this, apparently, was the way the company had done it. Greenberg had reportedly instigated the deal himself, with a phone call to Gen Re’s CEO. But, according to Spitzer and (later) the SEC, no real risk was transferred. The SEC would call the deal a “sham.”

  A few hours after Greenberg made his “foot fault” comment, Spitzer sent AIG a subpoena demanding documents relating to the Gen Re transaction. That evening, the New York attorney general happened to be the dinner speaker at a Goldman Sachs event. “Hank Greenberg should be very, very careful talking about foot faults,” he said. “Too many foot faults and you can lose the match. But more important, those aren’t foot faults.”

  Not surprisingly, the board was fast losing faith in Greenberg. Even before the Spitzer subpoena, the independent directors had hired Richard Beattie, a high-priced lawyer with Simpson Thacher & Bartlett, who was well known for advising boards faced with difficult situations. Beattie had several dinners with Greenberg, hoping to persuade him to retire. At one point, right around the time of the earnings announcement, Greenberg had agreed to step down. But the next day he told Beattie he had changed his mind.

  Once the Spitzer subpoena arrived, a special committee of directors began an internal investigation. AIG’s accounting firm, PricewaterhouseCoopers, was brought in to comb through the company’s books. “They were finding problems everywhere,” recalls a former AIG executive. It turned out that certain aspects of FP’s derivatives accounting were incorrect. Some reinsurance transactions had to be unwound. Deals that walked right up to the line—which PWC had once okayed—were now ruled out of bounds. And there was another problem: the company had promised to cooperate with Spitzer, and he wanted to depose Greenberg. The board wanted to know whether Greenberg was going to plead the Fifth Amendment. Greenberg said he wasn’t sure. How could the CEO take the Fifth and keep his job?

  The climactic board meeting took place on March 13, 2005. It lasted all day, with the directors discussing among themselves whether to fire Greenberg and Greenberg, calling in from his boat and his airplane, arguing that he should keep his job. He had an odd way of going about it, though. “You people don’t know what you’re doing,” he berated them. “You don’t even know how to spell insurance.”

  Toward the end of the meeting, the accountants from PWC told the board that it would no longer vouch for the firm’s books if Greenberg stayed as CEO. And that was that. Greenberg was allowed to stay on as chairman of the board, though that arrangement wouldn’t last long, either. His replacement as CEO was a bland AIG lifer named Martin Sullivan, who had spent his entire career on the insurance side of AIG. Although Sullivan had a tremendous amount of insurance experience, because of the way Greenberg ran the company he knew very little about AIG-FP.6

  Immediately after Greenberg’s departure, the rating agencies dropped AIG’s rating to double-A. Over the next few months, the int
ensity level inside the company was almost unbearable, as every subsidiary was turned inside out by swarms of accountants. “It was like a war zone,” says a former executive. In July AIG announced its restatement: the company would reduce its earnings by $4 billion covering the previous five years. In those five years, AIG had reported around $40 billion in profits; the new numbers lowered AIG’s profits by 10 percent. In other words, AIG didn’t really have to play games—$36 billion in profits would still have earned it plenty of respect. It’s just that Greenberg would have been seen as a mere mortal, instead of the great god of insurance.

  Alain Karaoglan, a Wall Street analyst who had followed AIG for years, wrote several searing reports in the wake of Greenberg’s resignation. One in particular stands out. After taking a close look at the Foreign Life unit—the same subsidiary that was always said to be one of the crown jewels—he concluded that there were significant gaps “… between statutory earnings reported in the 10-K and our summation of statutory account principle (SAP) earnings for the operating entities.” In other words, try as he might, he could not make the earnings that AIG had reported for Foreign Life add up.

  He also pointed out something that nobody had bothered to pay much attention to before. The rating agencies had consistently said that none of AIG’s operating subsidiaries would have merited triple-A ratings if they had been stand-alone companies. Only the guarantee from the parent company made them triple-A credits. Yet, he noted, “AIG, the parent, is just a holding company and its strength and only source of cash flow to bondholders and shareholders comes from its subsidiaries.” AIG’s vaunted triple-A, in other words, was a product of circular logic that broke down upon close inspection. As Karaoglan continued his analysis, he openly wondered whether the operating units deserved even a double-A rating.

  Then he wrote this: “[W]e were all to some degree complacent, and looked to some degree at the financials in a silo fashion and took comfort in the overall AIG whenever the silo could not stand on its own. In our view, we were all over-relying on Mr. Greenberg to sustain the company’s tremendous track record and ensure it was real. Now, with significant financial improprieties revealed by the company, we can no longer do that.”