All the Devils Are Here

by Bethany McLean; Joe Nocera

  By the time the Bank of America deal closed, the funds were in serious trouble. Investors had demanded half their money back from Enhanced Leverage, leaving Cioffi and Tannin with little choice but to wind it down. Tannin, at least, seemed to finally recognize that the jig was up. In late May, a Bear salesman announced good news: Tokio Marine wanted to invest $10 million in the High Grade fund. When Tannin heard the news, he asked for a meeting with Greg Quental, who was the head of Bear Stearns Asset Management’s hedge fund business. After the meeting, Quental announced that the Bear funds wouldn’t be taking any new investments.

  At the end of May, Cioffi e-mailed Spector, his longtime supporter at Bear. “Warren, I’m almost too embarrassed to call you,” he wrote. “I feel especially badly because you have been a big supporter of mine for so long… I know apologies are meaningless at this stage but I am sorry… Emotionally, I am obviously keeping a business as usual persona at work and on the job 24-7. I assure you of that. But it is very stressful and strange when it looks like one’s business is collapsing around him… we are running out of options.”

  More bad news was coming. A few weeks earlier, Bear had told investors in the Enhanced Leverage fund that it had lost 6.5 percent in April. But at a meeting on May 31, Bear Stearns’s pricing committee, which determined the funds’ returns by surveying how its counterparties were marking the securities, decided the fund had actually lost 18.97 percent in April. One key reason for the stunning change was Goldman’s low marks.

  According to the SEC, Cioffi tried to argue that his original marks were right. When he gave up, he wrote an e-mail to a member of the pricing committee: “There is no market… its [sic] all academic anyway −19 percent is doomsday.”

  Which, in fact, it was. While Bear could and did prevent its investors from taking their money out—a common tactic when hedge fund investors are all trying to exit at once—it was powerless to do anything about the repo lenders. “It’s like borrowing from the devil times three,” says one person who was there, speaking of the repo lenders. “They can come and say, ‘You pay me,’ and you can do nothing, nothing, nothing. There are no rules, and you have no ability to see where they’ve marked the same assets on their own books. It’s a grab.”

  On June 11, Cioffi wrote to Tannin and George Buxton, who worked in Bear’s private client services, “Right now we’re fighting the Battle of the Bulge with our repo lenders. So far so good but it is very tough and stressful…” He was trying to convince the lenders that if they grabbed the collateral and tried to sell it into a shaky market, everyone would get hurt. After all, the Bear team argued, they all had the same positions, and panicked selling would turn theoretical declines in the market value of the securities into hard cash losses.

  The men were also deeply frustrated. While Wall Street’s repo desks were demanding money, the trading desks at the same firms were refusing to reflect the value of the Bear team’s short positions. They were being squeezed from both sides. “The pressure was tremendous,” says one person who was there. “And everyone was scared.” On June 14, Bear held a meeting with the repo lenders to try to cut deals. At that meeting, according to House of Cards, William Cohan’s book about the fall of Bear Stearns, the Bear executives gave a presentation showing the exposure the rest of the Street had to the firm’s hedge funds. Overall, sixteen Wall Street firms had lent the funds $11.1 billion in the repo market. Among them were Citi, with nearly $1.9 billion outstanding, and Merrill Lynch, with $1.46 billion outstanding.

  A few days later, one firm broke from the pack. Merrill Lynch seized $850 million in collateral, which it said it would sell on the open market. Any chance of an orderly wind-down of the funds was now gone. It was a classic run on a bank—except that those racing to pull their money out weren’t depositors. They were bankers.

  When Merrill tried to sell the assets, it discovered that Cioffi had been right: nobody wanted to buy the collateral, at least not at the price that Merrill was valuing the securities. The firm largely abandoned the effort. J.P. Morgan and Deutsche Bank, which had followed Merrill’s lead, canceled their plans to sell assets, too. Here was the moment of truth: triple-A tranches of CDOs stuffed with subprime mortgages simply weren’t salable, not at a hundred cents on the dollar, and maybe not at any price. In fact, mortgage-backed securities weren’t salable, period. “All these guys grabbed for Bear’s mortgage-backed securities thinking they’d be able to write them up, not realizing they’d have to write them down,” says one person who was there. “All of a sudden, it became an internal witch hunt everywhere. How much of this do we own?”

  As Goldman’s Josh Birnbaum later wrote, “The BSAM situation changed everything. I felt that this mark-to-market event for CDO risk would begin a further unraveling in mortgage credit.” Goldman, which had covered its short position, quickly began to rebuild it.

  Although Bear itself did eventually put up $1.6 billion to try to save the High Grade fund, it wasn’t enough. On July 31, 2007, both funds filed for bankruptcy.

  But Wall Street was still too blind to see that the line between the Bear hedge funds—highly leveraged entities dependent on the repo market with big exposure to toxic subprime mortgages—and the firms themselves—highly leveraged entities dependent on the repo market with big exposure to toxic subprime mortgages—was a very thin one indeed. That lesson was yet to come.

  As the prices on triple-A-rated notes plunged in the early summer of 2007, the rating agencies continued to insist to the outside world that everything was just fine. At the beginning of the year, S&P had predicted that 2007 would bring “fewer ratings changes overall, and more upgrades than downgrades.” As the year went on and the skepticism about the validity of the ratings increased, the agencies claimed that they had run stress tests and scrubbed the numbers. Moody’s told Fortune, for instance, that its investment-grade-rated products were “designed to withstand losses that are materially higher than expectations.”

  The rating agencies were in the midst of a spectacularly profitable run. “The first half of 2007 was the strongest we had in five years,” Moody’s CEO Ray McDaniel would later say; its revenues had hit $1.2 billion over that period. Why? Because the Wall Street firms could all see the handwriting on the wall. With the ABX declining and triple-A tranches faltering, the CDO business was soon going to shut down. So Wall Street raced to shove as many CDOs out the door as it could; firms like Goldman wanted to get the bad paper off their own books onto someone else’s while there was still time. In that same six-month period, from January to June 2007, CDO issuance peaked at more than $180 billion. “[B]ankers are under enormous pressure to turn their warehouses into CDO notes,” Eric Kolchinsky, the Moody’s executive in charge of rating asset-backed CDOs, wrote in an August 2007 e-mail. Amazingly, the rating agencies continued to facilitate that effort by rating large chunks of these deals triple-A.

  Had the agencies noticed the increasing early payment defaults that had started in 2006? Of course. S&P and Moody’s had responded by increasing the amount of credit enhancement required to get investment-grade ratings on securities backed by subprime mortgages. But as the Senate Permanent Subcommittee on Investigations would later point out, neither agency went back to test old mortgage-backed securities or old CDOs using this new methodology. Thus, the old, flawed ratings continued to live on in portfolios all over Wall Street. Even worse, they were recycled into new synthetic CDOs, as old tranches were referenced in new securities. “Reevaluating existing RMBS securities with the revised model would likely have led to downgrades, angry issuers, and even angrier investors, so S&P didn’t do it,” said Senator Carl Levin, subcommittee chairman. Moody’s didn’t, either.

  Despite the optimistic glow the rating agency put on things to the outside world, there were plenty of people internally who feared the worst. In an e-mail exchange in early September 2006 among S&P employees, Richard Koch, a director in S&P’s structured products group, cited a BusinessWeek article on the bad lending practices in op
tion ARMs. “This is frightening. It wreaks [sic] of greed, unregulated brokers, and ‘not so prudent’ lenders… Hope our friends with large portfolios of these mortgages are preparing for the inevitable.” Six weeks later, Michael Gutierrez, another director in S&P’s structured products group, forwarded a Wall Street Journal story to several colleagues about how ever looser lending standards were leading to higher defaults. He wrote, “Pretty grim news as we suspected—note also the ‘mailing in the keys and walking away’ epidemic has begun—I think things are going to get mighty ugly next year!”

  “I smell class action!” responded a colleague.

  By February, S&P was having internal discussions about how to respond to the deteriorating value of mortgage-backed securities. “I talked to Tommy yesterday and he thinks that the ratings are not going to hold through 2007,” wrote Ernestine Warner, S&P’s head of global surveillance, to Peter D’Erchia, an S&P managing director. “He asked me to begin discussing taking rating actions earlier on the poor performing deals.” She continued, “I have been thinking about this for much of the night.”

  On March 18, one unnamed employee at S&P sent this in an e-mail: “To give you a confidential tidbit among friends the subprime brouhaha is reaching serious levels—tomorrow morning key members of the RMBS rating division are scheduled to make a presentation to Terry McGraw CEO of McGraw-Hill Companies and his executive committee on the entire subprime situation and how we rated the deals and are preparing to deal with the fallout (downgrades).”16 At Moody’s, the story was similar. The company’s own subsidiary, Economy.com, issued a prescient report in October 2006 called “Housing at the Tipping Point,” in which it reported, “Nearly twenty of the nation’s metro areas will experience a crash in house prices: a double-digit peak-to-trough decline.” A double-digit decline in housing prices is precisely what the rating agencies’ models said could never happen. In addition, big investors had started complaining that the ratings were flawed. At one point, Josh Anderson, who managed asset-backed securities at PIMCO, the giant bond manager, confronted Moody’s executive Mary Elizabeth Brennan. In an internal e-mail, Brennan reported, “PIMCO and others (he mentioned BlackRock and WAMCO) have previously been very vocal about their disagreements over Moody’s rating methodology.” She continued, “He cited several meetings they have had… questioning Moody’s rating methodologies and assumptions. He found the Moody’s analyst to be arrogant and gave the indication that ‘We’re smarter than you’…” Anderson went on to say, Brennan wrote, that “Moody’s doesn’t stand up to Wall Street… In the case of RMBS, its mistakes were ‘so obvious.’”

  And still the agencies continued to stamp their triple-As on mortgage-backed securities. The evidence didn’t seem to matter. In late December 2006, Moody’s analyst Debashish Chatterjee was shocked by his own graph of the number of mortgages at the top ten issuers that were more than sixty days delinquent. Fremont Investment & Loan, in particular, was drowning in them. “Holy cow—is this data correct? I just graphed it and Freemont [sic] is such an outlier!!” he wrote in an e-mail to colleagues. A month later, when S&P was rating a Goldman CDO that contained Fremont loans, the analyst on the deal asked a colleague, “Since Fremont collateral has been performing not so good, is there anything special I should be aware of.” The response: “No, we don’t treat their collateral any differently.” Both Moody’s and S&P rated five tranches of that offering triple-A; not surprisingly, two of the five were later downgraded to junk, according to analysis by the Senate Permanent Subcommittee on Investigations.

  It wasn’t until July 2007—the same month the Bear hedge funds collapsed—that the rating agencies made their first major move toward downgrading. E-mails imply that they had been considering such a move among themselves for months. It also appears that they were discussing it with at least some Wall Street firms as well. “It sounds like Moody’s is trying to figure out when to start downgrading, and how much damage they’re going to cause—they’re meeting with various investment banks,” a UBS banker had written back in May. A judge overseeing a lawsuit involving UBS would later find “probable cause to sustain the claim that UBS became privy to material non-public information regarding a pending change in Moody’s rating methodology.”

  Yet even in July, the rating agencies still weren’t ready to go all in and actually downgrade triple-A tranches. Instead, on July 10, 2007, S&P placed 612 tranches of securities backed by subprime mortgages on “review” for downgrade; almost immediately, Moody’s followed, placing 399 tranches on review. Both agencies made a great point of saying that the downgrades affected only a sliver of the mortgage-backed securities they had rated.

  Why had it taken so long? Sheer overwork played a part, as did paralysis. But it was also because the rating agencies feared the consequences of a widespread downgrade of mortgage-backed securities. With ratings so embedded in regulations, downgrades would force many buyers to sell. That forced selling, in turn, would put more pressure on prices, which would create a downward spiral that would be nearly impossible to reverse. With subprime mortgages, that situation was exacerbated a thousandfold, because the flawed ratings of residential mortgage-backed securities had been used to create countless CDOs—and synthetic CDOs. Downgrades of the underlying mortgage-backed securities could cause the CDOs to default even before any losses had shown themselves. The ripple effect was bound to be enormous.

  Later that day, S&P held a conference call for investors to discuss the pending downgrades. Most people were fairly polite. But one man on that conference call, a hedge fund manager named Steve Eisman, who had taken a big short position in mortgage-backed securities, was not.

  “Yeah, hi, I’d like to know why now?” Eisman began. “I mean, the news has been out on subprime now for many, many months. The delinquencies have been a disaster now for many, many months. Your ratings have been called into question for many, many months. I’d like to know why you’re making this move today when you—And why didn’t you do this many, many months ago?” S&P’s Tom Warrack, a managing director in the RMBS group, tried to break in. “We took action as soon as possible given the information at hand…” But Eisman wouldn’t be stopped. “I mean, I track this market every single day. The performance has been a disaster now for several months. I mean, it can’t be that all of a sudden, the performance has reached a level where you’ve woken up. I’d like to understand why now, when you could’ve made this move many, many months ago. I mean, the paper just deteriorates every single month like clockwork. I mean, you need to have a better answer than the one you just gave.”

  The next day, Mabel Yu, an analyst at Vanguard, told Mary Elizabeth Brennan at Moody’s that when Eisman started talking, “my phone was on mute but I jumped up and down and clapped my hands and screamed. He was the only one to say it, but all the investors were all feeling the same way.” Yu went on to tell Brennan that Vanguard had stopped buying mortgage-backed securities in early 2006 because they were less and less comfortable with the ratings.

  Although S&P and Moody’s wouldn’t actually begin downgrading CDOs until October, the party effectively ended that day in July. “[P] ut today in your calendar,” wrote Robert Morelli, who was in charge of the CDO business at UBS, to colleagues. When he was later asked what he meant by that, he explained, “to the day was essentially the beginning of the end of the CDO business.”

  A few weeks later, Moody’s Eric Kolchinsky forwarded some UBS research to colleagues. It showed that in a sample of 111 mezzanine asset-backed securities CDOs, the triple-B tranches could expect losses of 65 percent and that the losses would extend into the triple-A tranches. Kolchinsky quoted the UBS report to his colleagues: “This is horrible from a ratings and risk management point of view; perhaps the biggest credit risk management failure ever,” it said.

  On July 24, 2007, two weeks after the rating agencies made their first big downgrade move and one week before the bankruptcy of the Bear Stearns hedge funds, Countrywide announced its results for the
first half of the year. In a last, desperate grab for market share, Countrywide had waited until March 2007 to stop offering “piggyback” loans that allowed borrowers to purchase a home with no money down. As other, weaker correspondent lenders—those that made loans themselves but then sold their loans to bigger lenders—began to go under, Countrywide ramped up its business of buying loans. Since Countrywide was no longer entering into agreements to sell its loans before they were made or purchased, the company was bearing all the risk that the market would crack on its own books.

  The rot Mozilo had long insisted wouldn’t infect Countrywide had started to spread. Although the company announced a profitable quarter, investors were shocked to hear that its earnings had declined for the third quarter in a row on a year over year basis—and that delinquency rates on Countrywide’s subprime mortgages had more than doubled, to 23.7 percent, from less than 10 percent at the end of March. Delinquencies in prime mortgages—prime mortgages—also spiked. And the company revealed that it was taking several other hits, including $417 million worth of impairments, mostly due to declines in the value of home equity residuals, and another $293 million in losses in loans held on its balance sheet.

  “We are experiencing home price depreciation almost like never before, with the exception of the Great Depression,” said Mozilo on the company’s conference call that day.

  Morgan Stanley analyst Ken Posner was startled by the news. “That is just not a charge-off ratio one would expect for a—at least for an old-fashioned prime portfolio,” he said on the conference call.

  “Countrywide is a mortgage supermarket,” responded chief risk officer John McMurray. “So it is my belief that the portfolio that we have for the most part is going to be a good reference for what exists on a broader basis.”