But then, as he wound it up, he displayed where his heart really was: “By the way,” he wrote, “we must continue to grow our sales force and all other businesses that keep the top line increasing particularly in the origination channels.”
In late 2006, another meeting of mortgage executives was taking place, this one in Kauai, Hawaii. This was a gathering of Washington Mutual’s top producers. As part of the festivities, a handful of WaMu employees did a skit about a funeral for one of its competitors. At the podium, one employee solemnly read a note. “For this day, we have lost one of the true legends in our industry.” As he spoke, a coffin imprinted with a logo was carried out onto the stage by four pallbearers dressed in black, wearing black sunglasses. The logo read: COUNTRYWIDE.
“So many of us warned the dearly departed about the risky—some may say reckless—behavior they engaged in,” he continued. “Throwing money around like Paris Hilton and selling products they don’t really know or understand.” As the sounds of “Na Na, Na, Na, Hey Hey, Goodbye” filled the room, he added that there was a bright side to the passing of WaMu’s biggest rival: “[S] ome really scary and dangerous people won’t be on the street anymore.”
This was fiction, of course. At the time it took place, Countrywide was the biggest mortgage lender in the country. But the point was this: within the industry, it wasn’t any secret that Countrywide was out on the edge of the mortgage market, even if Mozilo himself didn’t want to believe it. Even WaMu, which was doing plenty of its own risky lending—enough to eventually bring it down—could see the excesses taking place at Countrywide.
It’s hard to know when the turning point took place at Countrywide. Risky loans were undoubtedly made on Kurland’s watch: he too pushed Countrywide’s market share ambitions. A shareholder lawsuit would later charge that Mozilo, Sambol, and Kurland were “principally responsible for [Countrywide’s] ‘culture change’ and concerted foray into leveraged and high risk lending practices.” According to this lawsuit, Kurland sold $192 million of stock from March 2004 to March 2008. But there were a few signals that lending wasn’t completely out of control. Eliot Spitzer had launched an investigation into whether Countrywide’s 2004 loans reflected racial bias. This was around the same time that Ameriquest was being investigated. In the end, Countrywide agreed to commit $3 million to consumer education—a far cry from the $325 million Ameriquest paid to settle the charges against it. One former executive says that Spitzer’s staff was crawling all over Countrywide; surely if they had discovered deeper problems, Spitzer would have come down harder on the company. (Countrywide cooperated with Spitzer, unlike J.P. Morgan, HSBC, and Wells Fargo, which took refuge in preemption.)
And at Countrywide, as with other mortgage originators, there had been a brief moment of sanity right before the Ameriquest settlement was announced. According to the Wall Street Journal, Countrywide was going to make it “tougher for borrowers to qualify for a 1 percent teaser rate on its option ARMs.” Internally, Kurland was pushing for that, according to a former executive; the company also issued a “no exceptions” policy in early 2006, meaning that there would be no more exceptions to underwriting policies. Besides, the government was going to issue that guidance on nontraditional loans, and Countrywide wanted to be on the right side of that. But as it became clear that any new guidance would have no teeth—and perhaps as Kurland lost power—the moment passed.
Once Kurland was officially out the door, Sambol began taking control of Countrywide. One of the first things he did was sideline some of the company’s governance structures, such as its executive risk committee, according to a former executive. Under Kurland, the protocol had always been to meet roughly a half dozen times a year. Under Sambol, it met once. Every meeting after that was canceled. “They devalued operational excellence and overvalued their own intellect,” says another former executive.
What’s more, no sooner had Kurland left than Sambol and Mozilo decided to switch regulators, shedding the OCC and the Fed for the OTS. “This move is one of the places where they made a terrible mistake,” says a former executive. Having the Fed and the OCC regulate the company gave it a bit of a halo effect that disappeared when it moved to the OTS. And really, insulting the Fed by cutting the regulatory cord was hardly a smart move.
By the end of 2006, Countrywide’s underwriting guidelines were “wider and more aggressive than they had ever been,” the SEC later charged. In a memo Mozilo sent to the board and all the top executives on December 7, 2006, he wrote that “subprime has evolved from a sector largely comprised of borrowers with impaired credit… to a sector offering very high leverage and reduced documentation.” And he noted the following shocking facts: In 2001, Countrywide’s maximum loan size in subprime was $400,000, with a maximum loan-to-value ratio of 90 percent (meaning a 10 percent down payment). You could do a stated-documentation loan only if you were self-employed. Countrywide did not have either interest-only loans or 80/20 loans in its product line. By 2006, however, subprime borrowers could get a loan up to $1 million. The maximum loan-to-value ratio was by then 100 percent. The only qualification for doing a stated-income loan was that you were a “wage earner.” Countrywide now offered interest-only loans to borrowers whose FICO scores were as low as 560, and 80/20 loans to borrowers with 580 FICO scores. As a result, 36 percent of Countrywide’s subprime originations in 2006 were done on a stated-documentation basis, versus just 13 percent in 2001. Twenty-three percent were interest-only, and 24 percent were 80/20 loans.11
To put it another way, it was hard to imagine anyone who wouldn’t qualify for a Countrywide subprime loan during the final throes of the housing bubble. In a lawsuit that would later be filed by the Mortgage Guaranty Insurance Corporation, which had insured many Countrywide loans, investigators went back and dug up details of some of the loans Countrywide had made during the subprime bubble. One loan Mortgage Guaranty investigated was for $360,000 made to a woman in Chicago who was supposedly earning $6,833 per month as an employee of an auto body shop. According to her loan application, the house she was purchasing was intended to be her primary residence. In truth, the woman was a part-time housekeeper who earned about $1,300 a month. She “posed as a front buyer to help her sister… and brother-in-law… acquire the home.” A few months after closing on the house, “[she] returned to her home in Poland because she was unable to find steady work in Chicago.”
Was this an example of a borrower pulling the wool over the eyes of the loan officer? Not exactly. “[The borrower] reported to MGIC that she disclosed her true employment, her actual income, and her intention to help her family purchase the property to the loan officer.” The loan officer told her she could “pose as a front buyer, obtain mortgage financing for her sister and brother-in-law, and avoid personal responsibility for the loan.” When the loan officer learned that she was a friend of the son of a man who owned an auto body shop, she “helped prepare a document” for the man to sign stating her employment and monthly income. Then she forged the man’s signature.
Another borrower was supposed to be a dairy foreman making $10,500 a month; he was really a milker at the dairy earning one-tenth that amount, and buying the house for his son rather than himself. The loan officer, according to the lawsuit, told him that he would be “lending your son your credit” and would not be responsible for the monthly payments. The borrower, who didn’t speak English, simply signed where the loan officer told him to. He got a $350,000 loan.
A “sales executive for Bay Area Sales and Marketing earning $8,700 a month” had actually been unemployed since 1989 and had no income. (And there was no such business as Bay Area Sales and Marketing.) She got a $398,050 refinancing. A house in Atlanta that had been appraised for $395,000 was worth no more than $277,000. A borrower’s tax return, claiming earnings of $17,661, was fraudulent, and his bank account was nonexistent. A borrower who claimed to be an account executive for “GNG Investments in Santa Clara, California”—another nonexistent firm—turned out to be
a janitor making $3,901.58 a month. She never made the $30,000 down payment Countrywide was claiming. She got a $600,000 house.
According to the Mortgage Guaranty lawsuit, “by about 2006, Countrywide’s internal risk assessors knew that in a substantial number of its stated-income loans—fully a third—borrowers overstated income by more than 50 percent. Countrywide also knew that many appraisers were overstating property values to drive originations by making loans appear less risky… Countrywide deliberately disregarded these and other signs of fraud in order to increase its market share.”
But those subprime loans weren’t the only thing that increased the risk drastically at Countrywide. In that schizophrenic e-mail Mozilo had sent back in May, he also wrote that “we must pay special attention to helocs [home equity loans] and pay options. With interest rates continuing to rise unabated helocs will become increasingly toxic…. As for pay options the Bank faces potential unexpected losses because higher rates will cause these loans to reset much earlier than anticipated and as a result causing mortgagors to default due to the substantial increase in their payments.”
In fact, Countrywide was originating huge numbers of pay option ARMs. In 2005 and 2006, Countrywide originated more than $160 billion worth of pay option ARMs—between 17 percent and 21 percent of its total loan originations, prime and subprime combined, according to a lawsuit later filed against the company. In 2007, with the market on the verge of collapse, Countrywide originated another $160 billion of these loans, according to Inside Mortgage Finance. The Los Angeles Times reported that Countrywide made one-quarter of all the option ARM loans in the country in 2007.
Although a high percentage of Countrywide’s pay option ARMs went to borrowers with high FICO scores—something Countrywide bragged about to its investors—that was a misleading statistic. The majority of the loans went to borrowers on a low- or no-documentation basis. And according to the Center for Responsible Lending, more than 80 percent of the option ARMs Countrywide originated in 2005 and 2006, totaling $138 billion, did not meet the new voluntary guidelines regulators had published in late 2006. In a letter to regulators, which was leaked to the Los Angeles Times, Countrywide admitted that it often judged whether borrowers could qualify for a loan based on the teaser rate, not the full rate, and that in the fourth quarter of 2006 about 60 percent of Countrywide’s adjustable-rate borrowers would not have qualified at the higher rate. (This appeared to contradict claims by Mozilo that the company’s policies required that borrowers be able to pay the higher rate.)
On July 10, Mozilo sent another e-mail to his top executives. “If I am reading these numbers correctly,” he wrote, “it appears to me that the loans (pay options) with neg am have a higher delinquency than our standard book of business. If this is the case, this is quite alarming, because of the very low payment requirements of a neg am loan.” He added in another e-mail, “I would like Gissinger and Hale to make certain that a letter, in BOLD TYPE, is included in every new pay option loan that clearly indicates the consequences of negative amort and encourage them to make full payment….”
And then there were home equity lines of credit, a product that was growing geometrically at Countrywide. As early as April 2005, John McMurray reported that the risk that home equity loans would default had doubled over the past year, mainly due to lack of documentation. That warning did nothing to slow the growth, nor tame the risk. Countrywide would later admit that a big chunk of its home equity lines resulted in the homeowner having debt that was close to 100 percent of the value of the property.
Back in the fall of 2006, with Sambol in charge and Countrywide’s market share hovering at just above 15 percent, the company put on a conference for investors. On the surface, at least, it was a high moment for the company. It would soon report 2006 revenues of $24.4 billion, up nearly $6 billion from 2005. Profits hit an all-time high of nearly $2.7 billion. Its ranking on the Fortune 500 rose from 122 to 91. So seemingly confident was the company in its financial strength that instead of conserving capital it announced a $2.5 billion stock buyback. In February 2007, Countrywide’s stock hit an all-time high of over $45 a share. What few at Countrywide seemed to understand was that it wasn’t just Countrywide’s customers who were assuming a great deal of risk. So was the company itself.
Like other mortgage originators, Countrywide kept the riskiest piece of a securitization, the residuals, on its own balance sheet. Kurland’s policy had been to presell subprime loans, the argument being that if you couldn’t sell the whole thing, then you shouldn’t make the loan. But a former executive says that changed. Another former executive recalls arguing to Drew Gissinger that these assets were risky and that the value at which Countrywide was booking them was inflated. Gissinger disagreed; these were high-quality assets, he said. “But that’s if everyone pays!” this executive responded. By the end of 2006, Countrywide had $2.8 billion worth of residuals on its balance sheet, representing about 15 percent of Countrywide’s equity. The company’s internal enterprise risk assessment map—a key risk report—was flashing orange.
Then, starting in 2005, Countrywide began to keep both pay option ARMs and a chunk of home equity loans—both the loans themselves and the residuals from home equity securitizations—on its balance sheet as well. In theory this made sense. Countrywide wasn’t just a mortgage shop, dependent on the vicissitudes of the mortgage market—it was a financial institution that could thrive in all markets. The rationale, once again, was that while there would be some delinquencies, the income stream from these loans would provide stability during tougher times. But, of course, that depended on the quality of the loans.
In the spring of 2005, Kurland argued that Countrywide was taking on too much balance sheet risk in home equity loans, according to the SEC. But the numbers just went higher. By the end of 2006, Countrywide had more than $20 billion worth of home equity loans on its books, almost double 2004’s level. And while Kurland had entered into hedges with Wall Street firms, offsetting the risk if the value of the residuals declined, those hedges were removed once he was pushed aside, according to one former executive. After all, by early 2007 they were in the money, and you could book a gain! “It wasn’t supposed to be about the gain,” says one former executive. “It was a hedge.”
Finally, Countrywide was putting pay option ARMs on its own balance sheet instead of selling them to Wall Street. By the end of 2006, Countrywide had $32.7 billion worth of pay option ARMs on its balance sheet, up from just $4.7 billion at the end of 2004. As Mozilo later wrote in an e-mail to Sambol and Sieracki, “We have no way, with any reasonable certainty, to assess the real risk of holding these loans on our balance sheet…. The bottom line is that we are flying blind on how these loans will perform in a stressed environment.” He began urging Sambol to sell the portfolio of option ARMs. But by that time, it was way too late.
There were some inside Countrywide who worried that the risks weren’t being adequately disclosed to investors. The SEC would later charge that, throughout 2006, McMurray “unsuccessfully lobbied to the financial reporting department that Countrywide disclose more information about its increasing credit risk, but these disclosures were not made.” In early 2007, McMurray provided Sambol and others with an outline of where it was likely to suffer losses.He asked that a version of the outline be included in the company’s year-end financial report. It wasn’t, according to the SEC. Later that year, he again argued that the company should disclose its widened underwriting guidelines to investors.
According to the SEC, Sieracki and Sambol made the decision not to include McMurray’s concerns about the underwriting guidelines in the company’s financial report. But it doesn’t seem like McMurray exactly laid his body across the tracks, either. He later said in a deposition that he was “comfortable” after discussing his issues with Anne McCallion, Countrywide’s deputy CFO. McCallion, for her part, said that “there were disclosures that were contained in the document that addressed the substance of his comments.”
>
But whether Countrywide was under a legal obligation to disclose more is almost beside the point. It was particularly important for a company like Countrywide that the market not get any nasty surprises, because Countrywide lived and died on the market’s confidence in it. Like all nonbank mortgage originators, Countrywide relied on cash from sales of its loans, and from selling equity and debt, to fund itself. Countrywide also relied on its ability to pledge its mortgages as collateral for loans in the overnight repo market. In fact, Countrywide was even more reliant on these funding sources because it also kept the rights to service the mortgages that it made, which it valued at $16.2 billion at the end of 2006. (Many other companies sold these rights.) Kurland had planned on this as a way to ensure that Countrywide could survive a market downturn caused by rising interest rates: the ongoing payments from servicing mortgages were supposed to provide a cushion in years when the company couldn’t make as many mortgages. But it meant that Countrywide got less cash in the door up front.
The risk of Countrywide’s dependence on the market could be mitigated if it were tightly managed, which explains why Kurland worried so incessantly about the operational aspects of the business. But the more loans and residuals that were put on Countrywide’s balance sheet, the harder the risk was to manage.
In other words, if the market ever got spooked about Countrywide’s health—if, say, investors began to question the value of the residuals or the loans on Countrywide’s balance sheet—and shut off the supply of cash, Countrywide could be in jeopardy.
All the Devils Are Here Page 31
