But you can’t keep a good investment banker from innovating, and in 2006 Bear Stearns came up with another bright idea, initiating a second scam. Whenever Bear Stearns learned of a loan default, instead of repurchasing the defective loan, it made a cash settlement with the originator at a discount, without informing the investors of either the breach or the cash settlement. So Bear kept the settlement cash, leaving the investors to discover the default later, which likely did not occur until after the investors’ contractual option to return the loan had expired. When a Bear Stearns manager specifically asked her boss if the policy was to make settlements with lenders without checking for violations of the representations and warranties that Bear Stearns had made to the investors, she confirmed that it was.12
At first Bear Stearns was overwhelmed by the flood of loan defaults, but it soon created a system to process them efficiently, generating $1.25 billion in settlements. Bear kept it all, and investors later learned that they owned securities backed by thousands of loans that Bear had officially listed as defaulted. The entire EPD process and the cash gains from the settlements were reported in detail to the most senior managers in the firm, so they were fully aware that investment agreements were being flouted and knew how bad the underlying loans were. Both the Bear Stearns auditor and legal counsel eventually insisted on stopping the one-sided settlements, although it seems that they continued for some months after the order.13
Underscoring that Bear managers well understood what they were selling, the deal manager on an August 2006 securitization called the deal a “shitbreather” and a “SACK OF SHIT”. In deposition, he said he intended those phrases as “a term of endearment”.14
All of the investment banks tried to maintain an appearance of propriety for a combination of legal and sales reasons. So, like all of them, Bear Stearns did not entirely ignore due diligence on the loans it bought, and as usual, had contracted with outside reviewing firms. But internal Bear Stearns e-mails mocked the low quality of the reviews, and managers consistently blocked proposals for tightened reviews. The head of one of the reviewing firms said in deposition that in a 2006 review, up to 65 percent of his rejection recommendations were ignored. Another reviewer said that about “75 percent of the loans that should have been rejected were still put in the pool and sold.”15
In 2006 Bear hired a third outside firm to check a sample of loans that had already been securitized. The review was subject to strict limitations: do not count occupancy violations (i.e., a declared prime residence not owner occupied); there could be no employment check; no credit report verifications; and no review of appraisals. Even with all those restrictions, 42.9 percent of the sampled loans were found to be in breach of the securitization conditions. Bear Stearns did not notify investors of those findings.16
By mid-2007 the impending collapse of the subprime housing bubble was becoming apparent. This initiated the third phase of the mortgage unit’s strategy. Instead of warning customers and investors, the unit went on an all-out drive to clear out its inventory—“a going out of business sale,” one manager called it. A rule against securitizing loans from suspended or terminated lenders was summarily dropped, without this fact being disclosed. Verschleiser railed about one $73 million batch of loans, three-quarters of which did not meet securitizing guidelines. He couldn’t understand “why any of these positions were not securitized . . . why were they dropped from deals and not securitized before their epd period.” And another senior trader demanded to know “why are we taking losses on 2nd lien loans from 2005 when they could have been securitized?????”17
In late 2007 Ambac managers became aware of the growing number of defaulted loans in the portfolios they had insured and requested a delivery of detailed loan files. Without telling Ambac, Bear Stearns hired one of its outside credit reviewers to look at the loans. The review found that 56 percent had material breaches. According to Ambac’s lawyers, Bear did not share this information with Ambac.18
Then came the fourth phase of Bear Stearns’s mortgage strategy—one practised on a far larger scale by others, as we shall see. Verschleiser realized that as the bubble ended, the resulting avalanche of loan defaults would have a catastrophic effect on Ambac, their insurer. Far from being a disaster, this was an enormous opportunity. Verschleiser realized that he could make a fortune—by betting on Ambac’s failure by shorting its stock. As he recounted in his 2007 self-evaluation:
[At] the end of October, while presenting to the risk committee on our business I told them that a few financial guarantors were vulnerable to potential writedowns in the CDO and MBS market and we should be short a multiple of 10 of the shorts I had put on. . . . In less than three weeks we made approximately $55 million on just those two trades.19
Bear Stearns took his advice, and by 2008 Verschleiser was consulting with other traders on other banks’ exposure to Ambac, so Bear Stearns could profit from its bad loans coming and going. In fact, the worse the loans, the more money Bear Stearns made. In mid-2007 Bear Stearns stock peaked at $159 a share, its all-time record.
But all good things must end. On 16 March 2008, Bear Stearns was facing bankruptcy, and its board agreed to sell the company to JPMorgan Chase for $2 a share (later revised to $10 after shareholder protests).
But if these guys were so good at obscuring the bad loans they were selling, why did Bear Stearns fail? At one level, the answer is simple: Bear Stearns ran out of money. It is very difficult to forecast with precision the end of a bubble, or the exact rate at which various market participants—banks, rating agencies, investors, insurers, executives—will catch on. So Bear Stearns got caught holding a lot of junk—bad loans, pieces of securitizations, stocks and bonds of other institutions also being decimated by the crisis—and couldn’t get rid of it fast enough, at high enough prices, because everybody else was waking up (and going down the drain) at the same time. Moreover, like all the investment banks, Bear Stearns was very heavily leveraged, dependent on huge quantities of dangerously short-term loans from money market funds and large banks. This funding needed to be rolled over weekly or even daily. When these funding sources sensed trouble, they stopped lending, and Bear Stearns ran out of cash very fast.
But why was the firm so dangerously exposed? Knowing the risks, knowing that a huge bubble would inevitably end, why had the firm continued to buy and hold so much junk, and why was it so reliant on such huge amounts of short-term funding? Well, the working-level people directly exploiting and profiting from the bubble didn’t have much incentive to end it, or warn anyone. As long as they could sell junk, they made money. When it started to go bad, many of them even made money by betting against it—by betting against specific securities, against indexes of mortgage-backed securities, and against firms likely to fail in the crash. And while they shared in the gains, all of the losses were someone else’s problem.
But that logic mainly applies to those who were most directly profiting from annual cash bonuses and could move to other firms. What about senior management and boards of directors, not just of Bear Stearns but of the others that collapsed—Lehman, Merrill Lynch, AIG, Citigroup, and even the lenders (WaMu, Wachovia, Countrywide, etc.)? In the event of failure, the CEOs and boards of these firms would unquestionably lose jobs and wealth not easily replaced. The answer to this question reveals a great deal, and we will consider it shortly. For now, let us continue our tour of Wall Street conduct.
We know more about Bear Stearns than we do about most other banks, due to the lack of government investigation and because the civil suits are being delayed. But sufficient material is available on other lenders to confirm the generality of bad behaviour. Some examples follow.
Goldman Sachs and GSAMP
GOLDMAN SACHS IS the other major bank for which there is a substantial internal record, thanks to documents obtained by the Senate Permanent Subcommittee on Investigations. The most interesting parts of Goldman’s behaviour are considered in the next chapter, because they pertain to what happene
d after the bubble ended. For now, I will simply provide one example to make the point that Goldman created junk like everyone else.
In the 15 October 2007 issue of Fortune magazine, Allan Sloan published his superb article “House of Junk”, which focused on a series of securities, the GSAMP Trust 2006-S-3, totalling $494 million out of the $44.5 billion in mortgage-backed securities that Goldman Sachs sold in 2006. The GSAMPs were issued in April 2006 (during Hank Paulson’s final months as CEO before becoming Treasury secretary), having been assembled from second mortgages sourced from among the worst subprime lenders, including Fremont and Long Beach. Since they were all second mortgages, the lender could not foreclose in the event of default. The average loan-to-value ratio of the pool was 99.29 percent, meaning that borrowers had essentially zero equity in the homes, and 58 percent of the loans had little or no documentation. Despite this, 93 percent of the securities were rated investment grade, and 68 percent were rated AAA, the highest possible rating, by both Moody’s and Standard & Poor’s, the two largest rating agencies. Yet by October 2007, 18 percent of the loans had already defaulted, and all of the securities had been severely downgraded.20
So it was junk. But did they know it was junk? They most certainly did; they started betting against it in late 2006, and by late 2007 they were already making net profits on their bets against mortgage securities. We’ll return to Goldman Sachs later in discussing how the financial sector handled the end of the bubble, the crisis, and the post-crisis environment. It’s interesting, and hasn’t been sufficiently publicized.
Morgan Stanley
MORGAN STANLEY WAS one of the leading global securitizers. In the first quarter of 2007 alone, MS created $44 billion of structured securities backed by mortgages and other assets. One modest deal sold to an unsophisticated institutional investor, the pension fund for Virgin Islands government employees, is a good illustration of banking ethics in the twenty-first century.21
The security in question was a particularly toxic thing called a synthetic CDO. A synthetic CDO is a kind of virtual, imitation CDO, not backed by actual loans or debts, but essentially a collection of side bets on other securities, constructed to track their performance. As with any bet, it takes two parties—someone betting that the securities will work, and someone else betting that they will fail.
The synthetic CDO in question here, Libertas, referenced, or made side bets on, some $1.2 billion of mostly mortgage-backed securities, a large share of them sourced from New Century, WMC, and Option One, all of them notoriously bad subprime lenders. The Virgin Islands pension fund bought $82 million worth of AAA-rated notes forming part of Libertas. The deal closed in late March 2007, and before the year was out, the securities were nearly worthless. But it gets better.
Morgan Stanley owned the short side of the entire deal—in other words, the people who created and sold these securities were betting that they would fail. So they did very well indeed when the pension fund’s notes defaulted, as they did within months. Since Morgan Stanley owned the short side, they kept the entire $82 million principal of the Virgin Islands pension investment. Nice work. You sell a deal, collect your sales commission, then you get to keep the customer’s entire investment when the securities fail.
That much is not in dispute. But, stunningly enough, it is not per se illegal to create and sell a security with the intention of profiting from its failure—a state of affairs that the investment banking industry is in no rush to publicize, much less change. So the question in the Virgin Islands lawsuit is whether Morgan Stanley knowingly misrepresented the quality of the securities. Here is the pension fund’s side of the story.
Morgan Stanley had long been New Century’s largest “warehouse” lender—supplying funds for New Century to assemble loans for securitization. As such, it carefully monitored conditions at the lender. We saw in the previous chapter that as the bubble ended and accounting problems surfaced, New Century rapidly declined into bankruptcy.
Morgan Stanley disclosed in the Libertas prospectus that New Century had been accused of trading and accounting violations, but did not mention the mounting claims for breaches of warranties. They also mentioned that “several published reports also speculated that [New Century] would seek bankruptcy protection or be liquidated.” Still, like all securitizers, they claimed they had vetted the loans and that they met standard quality guidelines.22
But Morgan Stanley knew a great deal more than it had disclosed. It had participated in a 6 March conference call with New Century and its creditors. After the call, Citigroup decided to invoke its default rights against New Century. About a week after the Libertas deal closed, Morgan Stanley seized $2.5 billion in New Century assets; New Century declared bankruptcy soon thereafter. The bankruptcy examiner later wrote: “[The] increasingly risky nature of New Century’s loan originations created a ticking time bomb that detonated in 2007.”23
The question in the lawsuit is whether Morgan Stanley deliberately withheld material information. But that was habitual for them. Like those of all the other securitizers, Morgan Stanley’s loans had been examined by Clayton Holdings, which, as usual, found that many of them violated even Morgan Stanley’s internal guidelines, and that many defective loans were securitized anyway.
In June 2010 Morgan Stanley agreed to pay $102 million to settle a lawsuit brought by the attorney general of Massachusetts. While not admitting wrongdoing, Morgan Stanley executed an “Assurance of Discontinuance” specifying a long list of improper acts and referencing a long list of past bad practices.
According to the settlement, when Morgan Stanley had been faced with a choice of maintaining its credit standards or continuing to source New Century loans, it chose to jettison standards. Morgan Stanley began to accept loans that didn’t comply with the Massachusetts “best interest” law, and progressively discarded its remaining internal quality rules. Even after Morgan Stanley declared New Century to be in default, it continued to provide funding for its mortgages—as long as the money was wired deal by deal to settlement accounts with availability only upon mortgage execution.24
And what about Option One? They were the H&R Block subsidiary, one of the “Worst of the Worst” subprime lenders. By the first part of 2007, delinquency rates on the loans from Option One that were used for Libertas were more than double the company’s earlier default rates even before the deal was closed. It does not appear that Morgan Stanley shared that information with the investors.25
During the bubble, Morgan Stanley had record profits, like everyone. As for the crisis, well, they survived, despite coming close to collapse in 2008. But Morgan Stanley would have done much better had it not been for one man, Howie Hubler, a senior trader whose erroneous bets on the mortgage market cost Morgan Stanley $9 billion.
But the nature of Mr Hubler’s bets is far more revealing than the size of the loss, particularly when compared to the rest of Morgan Stanley’s behaviour. Mr Hubler didn’t lose money because he innocently thought that mortgage securities were good things. Quite the contrary.
Like the people at Bear Stearns, Mr Hubler was anything but an obscure rogue trader. He ran a fifty-person group, and his decisions were reviewed by senior management. Mr Hubler realized by late 2004 that the housing market was a huge bubble, that it would burst, and that when that happened, thousands of mortgage-backed securities based on awful subprime loans would fail. Mr Hubler talked to his management about this, and they agreed with him.
Did Morgan Stanley then warn its customers? No. Did it stop selling tens of billions of dollars of crappy subprime mortgage-backed securities? No. Did it tighten its loan standards? No—indeed, as we have just seen, it lowered them. Did it warn the regulators? No. Did it stop financing the worst subprime lenders? No.
What Morgan Stanley did do, however, was give Howie Hubler permission to begin betting against subprime mortgage-backed securities, massively. Using credit default swaps—we’ll get to them—he placed enormous bets that very low-quality, but nonethel
ess highly rated, mortgage securities would fail.26
But there was a problem. The bubble lasted longer than anyone at Morgan Stanley predicted that it could. And as 2004 became 2005 and then 2006, maintaining Mr Hubler’s bets became expensive. But Mr Hubler was absolutely certain that the bubble would eventually burst, and he wanted above all to maintain his bets against those really awful subprime mortgage securities.
And so, with Morgan Stanley’s knowledge and approval, Howie made a huge mistake. In order to pay for his bets against the lowest-quality mortgage securities, he started writing insurance for other, supposedly higher-quality mortgage securities—securities that Mr Hubler thought would not default until much later than the really awful ones. But insurance on these higher-quality securities was much cheaper, so in order to sell enough insurance (to obtain enough premium income) to fund his bets against the obviously crappy securities, he needed to write insurance on a lot of them.
For a short time it worked, and in the first quarter of 2007 Morgan Stanley made $1 billion from Hubler’s strategy. But when the shit hit the fan, the supposedly higher-quality securities failed rapidly, too. Just as Morgan Stanley had underestimated the size and duration of the bubble, so too it had underestimated the severity of the collapse. Internal politics and/or sexism probably also interfered; an article published in New York magazine in April 2008 described power struggles and institutional sexism involving Hubler, other traders, John Mack (the CEO), and Zoe Cruz, Morgan Stanley’s highest-ranking female executive. Cruz was not a saint; she too endorsed the idea of secretly shorting the subprime market. But Cruz, to whom Hubler reported, apparently became alarmed about the potential risks of Hubler’s strategy.
Inside Job Page 11