Inside Job (16 page)

In early 2009 I also spoke with Stephen Roach, whom I had also known (slightly) for a long time. To his credit, Mr Roach had warned publicly about America’s unsustainable debt levels, and
of a coming recession. But he was careful never to blame his firm or industry; the culprit was Alan Greenspan, for keeping interest rates too low. When I asked Mr Roach if the structure of
financial sector compensation had contributed to the bubble, he said no, and argued that regulating compensation would be a bad idea. He didn’t mention Morgan Stanley’s betting against
the mortgage market either. Given that it cost his employer $9 billion, I find it difficult to believe that he didn’t know.

THE EVIDENCE SUGGESTS
that Citigroup behaved as unethically as Morgan Stanley and Bear Stearns did. But they weren’t as clever, or at least their
senior management wasn’t, so when the music stopped they were caught without a chair.

Although Citigroup had earlier applied fairly strict credit standards, when the bubble accelerated they held their noses and accepted
whatever the originators sent them.
By mid-2006 Richard Bowen, the recently promoted chief underwriter of Citigroup’s consumer division, grew seriously alarmed. He discovered that 60 percent of the loans that Citigroup was
buying from lenders failed to meet its own internal standards. He warned everyone around him, including senior management. But not only was nothing done, things actually got worse. By 2007 he said,
“defective mortgages . . . [were] over 80 percent of production.”
²⁷
He testified to the FCIC that he and other credit officers repeatedly
complained to the senior executives in the bank. In October 2007 Bowen wrote a very explicit e-mail, marked “URGENT—READ IMMEDIATELY,” about the failure and violation of
Citigroup’s internal controls, and their possible financial consequences. Bowen sent this e-mail message to four senior executives, including Citigroup’s CFO and also Robert Rubin, the
former Treasury secretary who was vice chairman of the board and chairman of the board’s executive committee. As a result, Bowen was demoted and his 220-person group reassigned, leaving him
with only two employees.

During the bubble, Citigroup purchased and resold huge volumes of mortgages, and also created and sold huge amounts of toxic mortgage securities. It sold many of these to the usual victims,
including Fannie, Freddie, and the Federal Housing Administration. But it also retained many of them, even though it pretended that it had not done so. Like the other securitizers, Citigroup made
use of highly deceptive accounting. In this case, it used a loophole in the accounting rules governing structured investment vehicles (SIVs). Actually it wasn’t a loophole at all, in the
sense of being an initially unintentional oversight later used for unanticipated purposes. Rather, it was a provision for which the banks had specifically lobbied hard (and won).

By placing its toxic mortgage securities in SIVs, Citigroup could take current profits and temporarily pretend that the securities weren’t on its own balance sheet—that is, until the
securities lost money, when it became necessary for Citigroup to pay up. In the meantime, of course, many people collected large bonuses, including Citigroup’s senior management. In the end
it turned out that Citigroup had ownership
or liability for over $50 billion of the stuff, which was why it had to be rescued by the US government. Late in the bubble,
Citigroup’s CDO unit did start to bet against the bubble, by creating synthetic CDOs that it could dishonestly sell to fools, thereby profiting by holding the short side of the bet. One such
synthetic CDO resulted in a civil fraud lawsuit filed by the SEC.
²⁸
But the profits thereby obtained were dwarfed by losses on Citigroup’s
holdings and the obligations it had to its SIVs.

Citigroup’s CEO during most of the bubble, Chuck Prince, was forced to resign in late 2007, replaced by Vikram Pandit, who remains CEO as of 2012. Rubin resigned in January 2009, forced
out when Citigroup became heavily dependent upon government rescue funds, which gave the government over 30 percent ownership.

UBS is one of Switzerland’s three big banks, and one of the largest banks in the world. It was both perpetrator and victim—some parts of UBS purchased huge quantities of mortgage
securities and lost billions on them, while others created and sold them to unsophisticated victims. In 2001—quite early—UBS created a hybrid CDO (part real, part synthetic) called
North Street 4. UBS sold it to a small, local German bank, HSH Nordbank.

The bank’s lawsuit against UBS alleges that contrary to representations, UBS and a subsequent acquisition, Dillon Read Capital Management (a small US investment bank), later used the
structure to dump poorly performing assets from its internal books. The last investment made on behalf of the investors was made by Dillon Read in February 2007. This investment was to go long on
an index of subprime mortgage CDOs (i.e., bet that they would perform well), so that Dillon Read could take the opposite position, betting
against
them. As of the date of the court filing,
HSH had lost half its investment.
²⁹

Essentially these same patterns of behaviour have been alleged against all the other major securitizers—Merrill Lynch, Deutsche Bank, Credit Suisse, Lehman. All of them sold securities
backed by high-risk residential mortgages sourced from the same universe of subprime originators. All of them had the same compensation practices.
And all the cases against
the big Wall Street firms we have met in this narrative come down to the simple proposition that no competent securitizer could have financed the loan originators, purchased tens or hundreds of
thousands of mortgages from them, structured and sold the securities, and made elaborate representations and warranties in their sales material without ever understanding the toxic nature of the
instruments they were selling.

In many cases, we have the additional evidence that they started betting against the market and even sometimes their own securities, while continuing to tell customers to buy. For example, late
in the bubble, traders at Merrill Lynch went to the extent of creating a new unit within Merrill to buy securities that the real market wouldn’t touch. Since that was obviously a losing game,
the traders split their bonuses with the fake purchasing group. In effect, the traders were financing bonuses by bribing their fellow traders, and all of them were ripping off the company’s
shareholders.

But there were two other major sets of players who were essential to the scam—the rating agencies and the insurers.

THE BUSINESS OF
rating debt securities was and remains an oligopoly, a quasi-cartel, of three firms—Moody’s (the largest); Standard &
Poor’s (S&P); and Fitch. For many years, they used their power to establish remarkable legal positions for themselves. The SEC recognized a limited number of Nationally Recognized
Statistical Ratings Organizations, and many government pension funds could only invest in securities that had high NRSRO ratings. The rating agencies avoided legal liability for their numerous
misjudgements by claiming that ratings were merely “opinions”, expressions of free speech protected by the First Amendment, and not subject to liability claims. When state legislatures
occasionally threatened to reduce their power or increase
their liability, the rating agencies threatened to stop rating bonds issued in those states.

For the last quarter century, the rating agencies have been paid by issuers, and as the investment banking industry itself consolidated, the banks increasingly called the tune, and the rating
agencies happily danced, all of them joined in corruption, cynicism, and exploitation. For the entire period of the bubble, and through 2007 as the bubble peaked and started to deflate, Wall Street
had pretended to ignore the impending crash. The rating agencies carried on issuing their triple-As with abandon.
³⁰

At both Moody’s and S&P, the volume of ratings processed on residential mortgage-backed securities (RMBS) doubled between 2004 and 2007. Mortgage-backed CDO ratings increased tenfold,
and each year the instruments grew far more complex. There was a similar boom in other bubble-related debt—instruments such as collateralized loan obligations (CLOs), auction-rate securities,
synthetic securities, and even more exotic objects, all of them routinely receiving extremely high ratings. The rating agencies became insanely profitable as a result. The relative stock price
gains for Moody’s outstripped even those of the best-performing financial services company, Goldman Sachs, by a factor of ten; during the bubble, Moody’s was the most profitable company
in the Fortune 500.

In the first week of July 2007, S&P rated 1,500 new mortgage-backed securities, or 300 per working day. It was an assembly line. In 2010 the former president of Clayton Holdings testified to
the FCIC that in 2007 he had approached all three major rating agencies, asking if they wanted the results of Clayton’s reviews of the loans being used by the securitizers. All three
declined.

By 2009 more than 90 percent of all 2006and 2007-vintage AAA subprime-backed securities were rated as “junk”.

Even more striking, however, was how the rating agencies treated the investment banks—their principal clients. Did they warn the world when Bear Stearns and Lehman became wildly
overleveraged,
when the industry shifted to unstable, ultra-short-term borrowing? No. Did they worry about the exposure of AIG and the other bond insurers? No. Even as the
crisis deepened, all of the securitizers and insurers, including all of those who failed or were rescued by the US government, continued to have high-investment-grade ratings—in several
cases, AAA.

Here is an excerpt from my interview with Jerry Fons, a former managing director at Moody’s who departed in 2007:

ONE LAST PART
of the system deserves comment: the protection racket, which in this context means selling insurance on mortgage-backed securities, making
investors feel even more confident that they were safe. This insurance came in two forms, bad and worse.

The less malignant form was literal insurance, sold by specialized
monoline insurance companies, the largest of which were MBIA and Ambac. They depended heavily on their
own AAA ratings, helpfully supplied as usual by the rating agencies. As everywhere, their salespeople and executives made lots of money during the bubble, which they kept even after securities
started to fail and the companies faced an avalanche of claims; both companies suffered disastrously. But at least they had previously been known principally for writing actual insurance, purchased
by the actual owners of actual bonds. For true insanity, one needs to look at the business of selling derivatives called credit default swaps (CDSs), the most famous practitioner of which was, of
course, AIG.

Credit default swaps are pure gambling, and they differ from insurance in extremely important ways. You can use CDSs to bet that any security will fail, and you can make your bets as large as
you want. The reverse is true as well: if you’re crazy enough, you can
sell
as much “protection” as you want, far beyond the real value of the securities in question. AIG
sold around $500 billion of it—and it didn’t work out too well.

CDSs were (and remain) particularly dangerous for several reasons. First, courtesy of the Commodity Futures Modernization Act of 2000, the law championed by Larry Summers, Robert Rubin, and Alan
Greenspan, CDSs were totally unregulated and extremely opaque.
Nobody
knew the total size and distribution of CDS ownership or risk, and the government did not have the legal right to
control it. Second, CDSs generated dangerous incentives—if you owned enough of them, then you had powerful incentives to
cause
something to fail, and also to remain silent about risks
in the financial system.

CDSs also provided the illusion, and sometimes the reality, of total protection against even the riskiest, most dangerous financial behaviour. Real insurance policies guard against
irresponsibility and fraud by having deductibles, policy limits, higher premiums for people with bad records, and other constraints. You cannot buy an insurance policy on your house for twenty
times its real value, especially if your house has conveniently burned down five times in the last decade. You also
cannot buy a home insurance policy if you don’t own
a home. If you have five convictions for drunk driving, you may not be able to buy car insurance at all, and if you can, it will be very expensive.

Unlike real insurance, CDSs had no deductibles or policy limits, and placed no constraints on buyers. As long as you had the money, you could buy as many of them as you wanted, even on
securities that you did not own—in other words, you could bet that a security, and/or the company issuing it, would fail. If you were the creator and issuer of such securities, you could
therefore profit by creating and selling junk, and then betting against it.