by Sheila Bair
We were able to get the OCC to agree to apply the same “fully indexed rate” standard we had used in the NTM guidance. I argued, successfully, that it made no sense and would be confusing to banks as well as consumers to have differing standards, depending on whether the bank was originating a nontraditional mortgage or a subprime loan. But we had a particularly hard fight with Randy Kroszner at the Fed over prepayment penalties. I felt that the penalties were anticompetitive and arguably discriminatory, as they were not used with prime borrowers. From a safety and soundness perspective, we did not want hybrid ARM borrowers locked out of refinancing as home prices continued to decline. In the end, we compromised25 on language providing that prepayment penalties should expire no later than sixty days before the interest rate reset.
Unfortunately, it was too late for the guidance to have a major impact. For one thing, much of the damage had already been done. By the second quarter of 2007, subprime loans outstanding totaled $1.3 trillion. For another, both the subprime and NTM guidance documents were enforced unevenly by the regulators, particularly the OTS and state regulators that had authority over nonbank lenders. The FDIC was the primary regulator for only one major subprime lender, Fremont Investment & Loan, which we chased out of the subprime business in February 2007, well before the guidance was finalized. By doing so, we averted what would have been a costly failure. But instead of cracking down on thrifts doing high-cost mortgage lending, the OTS let a number of them grow their portfolios of subprime and NTM loans. By the summer of 2007, the major nonbank mortgage lenders such as New Century and Ameriquest were starting to fail. Unbelievably, OTS was allowing its major thrifts to pick up a lot of their business. Insured thrifts actually grew26 their mortgage loan balances from $727 billion at the end of 2006 to $795 billion by the third quarter of 2007. All of the high-risk mortgage lenders regulated by the OTS eventually failed or were acquired. Their losses were massive. None survived, and the OTS itself was abolished by Congress.
The subprime lending abuses could have been avoided if the Federal Reserve Board had simply used the authority it had since 1994 under the Home Ownership Equity Protection Act (HOEPA) to promulgate mortgage-lending standards across the board. The Fed was the only government agency with the authority to prescribe mortgage-lending standards for banks and nonbanks. As the Financial Crisis Inquiry Commission (FCIC) concluded in its 2011 report:
There was an explosion27 in risky subprime lending and securitization, an unsustainable rise in housing prices, widespread reports of egregious and predatory lending practices, dramatic increases in household mortgage debt. . . . Yet there was pervasive permissiveness; little meaningful action was taken to quell the threats in a timely manner.
The prime example is the Federal Reserve’s pivotal failure to stem the flow of toxic mortgages, which it could have done by setting prudent mortgage-lending standards. The Federal Reserve was the one entity empowered to do so and it did not.
In March 2007, I testified strongly in favor of the Fed issuing an antipredatory lending regulation under HOEPA; it was a step that consumer advocates such as Martin Eakes had been pushing for years. I could barely contain my anger when the Fed’s witness countered with a go-slow approach. Disregarding the pronounced deterioration of the subprime market, she argued that the Fed needed to do a “careful review”28 and it was concerned about “constraining responsible credit.” In July 2008, Chairman Bernanke would finally direct the reluctant Fed staff to promulgate HOEPA regulations, which would take effect at the beginning of 2009. By that time, though, the damage was done. Ben publicly acknowledged that their failure to act earlier had been a key shortcoming in the Fed’s handling of the crisis. Why did the Fed delay? As Fed General Counsel Scott Alvarez put it, “The mind-set was29 that there should be no regulation; the market should take care of policing, unless there already is an identified problem. . . . We were in the reactive mode because that’s what the mind-set was of the ’90s and the early 2000s.”
CHAPTER 5
Subprime Is “Contained”
The resistance we encountered from other regulators in tightening up subprime lending standards was symptomatic of their early tendency to discount the severity of the potential problems. Even into 2008, both the Fed and Treasury (of which the OCC and OTS were a part) downplayed potential systemic risks from the subprime debacle.
At the FDIC, we were not so sanguine. Richard Brown, our chief economist, and Christopher Newbury, the head of our risk analysis research unit, had been monitoring the performance of housing generally and subprime loans in particular. At a briefing for me in October 2006, they presented troubling data. The performance of subprime mortgages was deteriorating markedly. At the same time, insured banks’ purchase of securities backed by these loans was increasing.
Unfortunately, their analysis was hampered by a lack of data: our information systems were limited to loans FDIC-insured banks held on their balance sheets. Since most of the subprime and other high-risk loans were being sold into Wall Street securitizations by both bank and nonbank mortgage originators, it was necessary to purchase information about those loans from private vendors. I authorized purchase of a private database, and once we had the data, our analysis of it confirmed the staff’s worst fears. The loans were a true parade of horribles: lack of documented income, little if any down payments, steep interest rate adjustments, abusive prepayment penalties, and mortgage payments that frequently exceeded 50 percent of the borrower’s gross income.
I couldn’t believe it. When I had served at the Treasury Department in 2001–2002, I had worked with the late Ned Gramlich on predatory lending practices in lower-income neighborhoods. At the time, Ned was a governor with the Federal Reserve Board and had responsibility for consumer issues. Consumer groups were reporting increasing instances of unregulated, nonbank mortgage brokers entering lower-income neighborhoods and “push marketing” mortgages with steep payment resets, negative amortization, and exorbitant prepayment penalties. Those were not “affordability” loans. Rather, the brokers frequently targeted existing home owners who had built equity in their homes, convincing them to pull cash out of their houses by refinancing their safe thirty-year fixed mortgages into complex subprime mortgages. The brokers were not banks and thus fell outside the lending standards applicable to insured institutions. Some were affiliated with banks, but the Fed had not used its authority over bank affiliates to examine the brokers for abusive practices, even though Ned had pushed Chairman Greenspan to do so.
In 2000, the Treasury Department and the Department of Housing and Urban Development (HUD) had issued a report recommending stronger lending standards for both bank and nonbank mortgage originators. The report suggested that either Congress enact a law to address mortgage-lending abuses for both banks and nonbanks or the Fed use the authority it had under HOEPA to establish marketwide lending standards. Notwithstanding Ned’s concerns, the Fed was disinclined to use its rule-making authority, and the industry had successfully stopped antipredatory lending proposals on Capitol Hill. In 2005, Congressmen Barney Frank and Spencer Bachus (R–Ala.) tried to put together a bipartisan effort to establish national lending standards. However, the effort met30 stiff opposition from the industry, which complained to the Republican leadership. Bachus was forced31 to stop negotiating with Frank under pressure from the House GOP leadership. Similarly, Senator Paul Sarbanes’s efforts to pass a national antipredatory lending law were stymied by industry lobbying efforts.
In the face of federal inaction, a number of state legislatures had enacted antipredatory lending laws that were helping somewhat. However, the OTS gave the thrifts that it regulated “field preemption32,” meaning that its regulation of mortgage lending prevailed over state laws and those laws, for the most part, could be ignored by the national thrifts. When I served at Treasury, the OCC was considering giving the national banks it regulated the same broad preemption. I received briefings from Julie Williams, the OCC’s general counsel, on its plans, as well
as from the comptroller, Jerry Hawke. As I was assistant secretary for financial institutions, they were obliged to consult me on banking policy issues. (By contrast, there were strict firewalls between the OCC and the Treasury Department regarding supervisory matters for particular institutions, as was appropriate.)
I had no authority other than to give the OCC my views, but that I did, and I didn’t mince words. I thought that preempting state consumer mortgage-lending laws was a singularly bad idea, particularly since the OCC had failed to promulgate any rules of its own to address the abuses we were seeing. The states were trying to protect their citizenry against a growing array of harmful lending practices. Unless the OCC was going to promulgate standards providing the same level of protection, I didn’t think it should be getting in the states’ way. In taking that position, I was given strong support by two of my career staff Treasury advisers, Edward DeMarco and Mario Ugoletti. We were highly skeptical of the OCC’s objectives and suspected that by expanding the scope of state preemption, the OCC hoped that large, state-regulated banks such as JPMorgan Chase would “flip” their charters and become national banks. The OCC backed off its proposal while I was in office, but in 2003, after I left the Treasury, it moved ahead. Sure enough, soon thereafter, JPMorgan Chase switched from being chartered by New York State to being OCC-regulated.
Ned and I worked together to convince the major mortgage lenders to agree to a set of best practices that would address the steep payment resets and lack of affordability. But the agreement was voluntary, and it did not hold. By 2006, practices that Ned and I had viewed as predatory in 2001 had become mainstream among most major mortgage lenders. How could things have deteriorated so quickly in five years?
In a word, securitization.
In its most basic form, the securitization process involves an issuer—typically a major financial institution—that accumulates a large volume of residential mortgages. The issuer might originate the mortgages itself, or it might obtain them from other lenders or independent mortgage brokers. Working with a Wall Street investment bank, the issuer packages the mortgages together into “pools” and divides the right to the cash flows of those mortgages into securities that are sold to investors, typically institutional investors such as pension funds, mutual funds, hedge funds, and insurance companies, as well as Fannie Mae and Freddie Mac (more about them later). The securities are sold from different “tranches.” That simply means that the securities are grouped into different priorities for payment of the cash received from the mortgage payments. The top, or more senior, securities must be paid in full before the lower, or more junior, securities receive any payments. As a result, if some of the mortgages in the pool default, investors holding the junior tranches will suffer losses first, protecting investors holding the senior tranches.
Here is a highly simplified example. Issuer X pools $1 billion worth of mortgages into a securitization and divides the pool into three tranches. To keep the math simple, let’s say that if all the mortgages perform, they will produce $100 million a year in mortgage payments. Investors in the top tranche of the securitization buy securities that promise $80 million of the cash flows each year, and under the terms of the securitization, they must be paid before any other investors. Investors in the second tranche, commonly called the “mezzanine” tranche, buy securities promising $10 million of the annual cash flows. They are paid only after the top tranche has received all of their expected proceeds. Investors in the bottom, or “equity,” tranche hold an interest in the remaining $10 million in cash flows. However, they are paid last. Investors in the mezzanine and equity tranches will pay less for their shares, as they have more risk than the top tranche. If any of the loans in the pool starts to default, the cash flows to the equity- and mezzanine-tranche investors are reduced before the upper-tranche investors are impacted. So it would seem that the upper tranche has substantial protection: loans representing more than 20 percent of the cash flows would have to default before their payments would be reduced. And prior to 200633, mortgage loans had historically very low default rates, less than 2 percent.
That extra level of protection for the top or senior tranches, known as “overcollateralization,” was the primary reason the ratings agencies routinely gave them triple-A ratings. Unfortunately, neither the investors nor the ratings agencies looked down into the pools to adequately analyze the quality of the individual loans in them. (And in fairness, the SEC did not require detailed loan-level disclosure and enough time for investors to analyze the disclosures before they invested, a problem it is now trying to correct.) If they had, they would have seen what we saw when we bought our data: little income documentation, high debt-to-income ratios, and steep, unaffordable payment resets. The “affordability” of the loans was determined based on borrowers’ ability to refinance or flip the property, not on a documented income capacity to pay. Thus, as the housing market turned and home prices started to decline, massive defaults could be expected.
The ratings agencies and investors did not do their homework. But what about the financial institutions originating the loans? Why didn’t they do a better job? Well, ask yourself: if you ran a business where you could sell a product and be paid up front, while suffering no losses regardless of how defective the product might be, how would that impact your behavior? From a purely economic standpoint, you would generate as much volume as possible to maximize your income. And that is exactly what happened.
Prior to securitization, mortgage lending was dominated by banks and thrifts, which would use customers’ deposits to make and hold mortgages to their customers. They were careful about loan quality because if a mortgage went bad, the loss was theirs. But with the advent of securitization, the funding came from investors, not depositors, so a bank—or a balance sheet—wasn’t required to make a mortgage loan. Stand-alone, nonbank mortgage lenders, such as the defunct Option One and New Century, churned out mortgages funded by immediate sale of the loans into securitizations. The lenders were virtually unregulated. Poorly trained and equally unsupervised mortgage brokers popped up all over the country, particularly in “hot” housing areas such as Florida, Nevada, and California. The brokers would originate the mortgages and sell them to a nonbank lender. The lender would get a short-term loan—frequently from a large national bank or thrift—to fund the mortgage for the home owner. The lender would then pay the loan back once the mortgage was sold to a securitization set up by a Wall Street investment bank and immediately pocket the rest of the securitization proceeds as profits.
The short-term loans made by commercial banks and thrifts were called “warehouse” loans, and they were permitted by banking regulators because the institutions held risk for a very short time period before the loans were repaid from securitization proceeds. In that indirect way, a number of large national banks and thrifts helped fuel the subprime crisis, even though they did not originate the loans. The FDIC pushed to have the subprime and NTM guidance apply to loans that banks and thrifts funded through warehouse loans, but the idea was strongly opposed by the industry and resisted by the other regulators.
Some FDIC-insured banks and thrifts also originated their own subprime and NTM loans, which were of somewhat higher quality than those purchased from independent mortgage brokers. But there again, economic incentives to assure good-quality underwriting were weak because the banks were selling the loans and passing on the risk of future default to investors through securitizations. Regulators and accountants also gave banks incentives to securitize loans. If they kept the loans in their own portfolios, they would have to hold capital and reserves against them to protect against potential losses. If, on the other hand, they sold them into securitizations, the regulators and accountants assumed that the risk had been transferred, so there were no capital and reserve requirements.
However, at the FDIC, we were concerned that this risk transfer was illusory. When large banks and thrifts sold loans into securitization pools, the contracts typically included “
reps and warranties,” which obliged them to buy back a defaulting loan if it did not meet the lending standards specified in the contract. As competitive pressures to relax lending standards intensified, we had no confidence that the quality of bank and thrift securitized loans would be sufficient to withstand demands by investors to “put back” bad loans. In addition, even if the loans met the contractually specified standards, we suspected that the major banks and thrifts would feel obliged to buy back defaulting loans to maintain their reputations with investors. Our fears were on target. So far, the major banks34 have bought back billions of securitized loans and, by their own estimates, face another $72 billion in litigation exposure.
Our staff was also concerned about insured banks’ risk in holding the equity tranches, also known as “residuals” of the securitizations that they sponsored. Remember the 10 percent equity tranche in the above example? Frequently, a sponsoring bank would keep that tranche of securities and sell it into another securitization-type structure called a collateralized debt obligation (CDO). This is how it would work: a Wall Street firm would buy the equity tranches and pool them into a securitization similar to the structure it used for mortgages. That is, it would “retranche” the securities using the same technique of overcollateralization. It would then slice up the pool of those tranches, creating a new top tranche that had a priority claim on the expected cash flows, with the lower tranches having the right to any cash flows that were left after the top tranche was paid. There again, the ratings agencies would hand out triple-A ratings to the top tranche, believing that there was a very low probability that all of the equity tranches would go bad at the same time. Of course, that was a fallacious assumption—the equity tranches were the most toxic pieces of larger pools of very toxic mortgages, and they almost all did go bad.