by Sheila Bair
Adding to this witch’s brew of poisonous assets was the credit default swap (CDS) market. Credit default swaps are a form of protection that investors and others can buy against the risk of default on a security they own. Investors in CDOs would “wrap” them in CDS protection provided by an insurance company or derivatives dealer. (American International Group—AIG—sold a lot of that kind of protection, eventually leading to its failure and need for a bailout.)
The CDO market has been all but wiped out. Some big banks or their affiliates—Citigroup being a prime example—had a large exposure to the CDO market. And there again, the regulatory and accounting treatment of those products provided all the wrong incentives. If banks kept the equity tranches on their balance sheets, they had to hold a substantial amount of capital and reserves against them. If they sold them to a Wall Street firm to be put into a CDO and then bought back the triple-A-rated portion, the capital requirement was quite low.
Magnifying the losses was the fact that Wall Street firms created “synthetic” CDOs, which allowed big financial institutions and their trading partners to make speculative bets on how certain CDOs would perform without actually owning them. Similarly, hedge funds and other speculators could buy CDS protection against the default of mortgage-backed investments, without actually owning them. It was like a game of fantasy football, with speculators tracking and wagering money on the performance of synthetic—that is, pretend—mortgage investments. That created trillions of dollars in speculative trading, many multiples of the size of the underlying subprime mortgage market. That is why hundreds of billions of dollars’ worth of mortgage losses translated into trillions of dollars of trading losses—losses that brought our financial system to its knees and caused the worst recession since the Great Depression. The accounting rules allowed many of the speculative investments to be held off balance sheet, making their risks invisible to regulators and market analysts until they started producing losses. I believe that speculative use of CDSs should be banned, given the damage they caused to our economy.
Another source of indirect exposure to the subprime mortgage market was reflected in FDIC-insured banks’ and thrifts’ huge holdings of securities issued or guaranteed by Fannie Mae and Freddie Mac. Fannie and Freddie, known as government-sponsored enterprises (GSEs), were created by Congress decades ago to provide support for housing finance. Fannie, like the FDIC, was a Depression-era New Deal agency. As envisioned by Congress, Fannie and Freddie were supposed to buy good-quality mortgages from mortgage lenders, package them into pools, and sell them to investors. To entice investment to support housing finance, Fannie and Freddie would guarantee continued principal and interest payments on those securities, even if the mortgages backing them went sour. Since Fannie and Freddie retained the risk that the mortgages might default, they would charge a guarantee fee from the banks and other lenders that originated and sold them the loans. Congress also required that in carrying out their mission, Fannie and Freddie support affordable housing for low- and moderate-income borrowers.
In the late 1990s and early 2000s, some people at Fannie and Freddie got the idea that they could make a lot of money by issuing bonds at very cheap interest rates (because investors assumed, given their government charter, that they were implicitly backed by the government) and using that money to invest in subprime mortgage-backed securities (MBSs) from Wall Street firms. The securities had very high rates of return (remember those 7% to 9% interest rates on 2/28s and 3/27s). By using that strategy, Fannie and Freddie operated like giant hedge funds, raising money cheaply from the debt markets (because bond investors assumed that they were backed by taxpayers) and using that cheap money to buy high-yielding mortgage-backed securities from Wall Street. For a while, Fannie and Freddie made money like crazy, paying their executives and shareholders huge bonuses and dividends. Indeed, in 2005, they gobbled up a full one-third of such securities from Wall Street. The Department of Housing and Urban Development actually encouraged them to do this by counting their investments in these toxic securities as credit in meeting “affordable housing” goals.
Insured banks and thrifts happily bought the debt the GSEs issued that supported the investment strategy, again with encouragement from the regulatory structure. That is because for decades, bank regulations viewed GSE debt as being nearly as safe as Treasury debt and thus required that banks hold very little capital against those securities. By giving banks favorable capital treatment on GSE debt, bank regulations encouraged insured institutions to fund them and reinforced the notion that the GSEs had an implicit backstop from the federal government. Of course, their investments in subprime securitizations eventually went sour, generating huge losses and forcing the GSEs into government conservatorship, where they still operate.
To sum up, between 2001 and 2006, the securitization market had exploded. Without adequate regulatory oversight, all of the economic incentives were to generate loan volume regardless of loan quality. There were no effective regulatory requirements that those who originated and securitized these loans hold any of the risk that the loans might default later on. Investors were holding the ultimate risk, but they did not do their own due diligence, nor did the ratings agencies, on which investors relied. The majority of the toxic subprime and NTM loans was originated by nonbank originators, though large banks and thrifts played a significant role. Some of the large banking organizations, most notably Citigroup, had substantial subprime operations. In addition, many large banks and thrifts provided short-term “warehouse” financing to nonbank mortgage originators making toxic subprime loans. Finally, the GSEs played a major role not by directly buying and guaranteeing subprime mortgages but after the fact, by gobbling up the senior-tranche subprime securities packaged and sold by Wall Street investment houses. Though greed was the primary motivation for the subprime craze, the government played a role by giving favorable capital and accounting treatment to banks that securitized their loans and also by giving the GSEs credit toward their affordable housing goals.
In the second quarter of 2006, when I assumed the chairmanship of the FDIC, the direct exposure of insured banks to privately issued securitization interests was not substantial: $252 billion, or about 2.2 percent of total assets. There was a temptation to take comfort from that relatively low number. Fortunately, the FDIC had had extensive experience with securitization; indeed, it had pioneered using securitization as a way to dispose of toxic assets inherited from failed banks and thrifts in the 1980s. Our staff knew all too well about put-back risk, as well as the larger dangers of steep home price declines when the housing bubble popped and the unsustainable mortgages started going bad. I think other regulators kept believing that the subprime problem was going to hit primarily the nonbank lenders that had originated most of the toxic loans; hence their public comments that the problem would be “contained.” Certainly, the specialty subprime lenders were the first to go; not a one has survived on a stand-alone basis—they have either been purchased by another financial institution or have gone bankrupt. But even if their direct exposure to subprime loans was limited, insured banks had huge exposure to residential real estate in general. They held more than $4 trillion35 in residential real estate–backed loans, representing more than 36 percent of bank assets. Our economists understood that substantial declines in the collateral backing those loans could have severe ramifications for the banks we insured.
So we had very good reasons for aggressively pushing stronger lending standards, even though I know the other regulators resented our advocacy. They, not the FDIC, were the primary regulators of the big players in the mortgage market, and they felt we were meddling in their business. Tensions became even more pronounced when we started advocating for banks and thrifts to restructure subprime mortgages. The truth is, by the time we finalized the guidance strengthening subprime lending standards, most of the damage had been done already. By mid-200736, there were $1.8 trillion of mortgages securitized by Wall Street outstanding. More than 7.5
million U.S. households held subprime loans. Strengthening lending standards would help only prospectively. What about all the bad loans that had already been made?
There again, the FDIC staff were on top of the analysis. Having long used securitization to dispose of failed bank assets, the FDIC had some of the top experts in the field. They included James Wigand, who ran most of our failed-bank resolutions, and one of his deputies, George Alexander. Also on the team was Michael Krimminger, who served as my top legal adviser throughout the crisis. A senior FDIC attorney, Mike was one of the foremost legal experts in the area of securitizations and bank resolutions. Working with data provided by Rich Brown and his team, we decided that the toxic loans need to be restructured in scale to avoid a foreclosure crisis that would put severe downward pressure on home values and throw millions of American families out of their homes.
CHAPTER 6
Stepping over a Dollar to Pick Up a Nickel: Helping Home Owners, Round One
Loan restructuring, also known as “modification,” is a time-tested tool used in the banking industry to minimize losses when a borrower runs into trouble. It is almost always the case that foreclosing on a distressed borrower—be it a business or a consumer—and seizing and liquidating the collateral will generate heavy losses for the lender. Therefore, if the terms of the loan can be changed, or modified, to make it affordable so that the borrower can resume making payments, the lender will usually mitigate its losses and realize substantially greater recoveries. Think about it. Losses on foreclosed property sales can typically approach 40 to 50 percent of unpaid principal balances. (The losses can be much higher if the distressed borrower owes significantly more than the property is worth, as is now the case with more than 10 million home owners.) That means the lender has a lot of leeway to make a significant payment reduction before it will become more profitable for it to foreclose on the loan. Why not try modification first? After all, if the borrower defaults again on the loan even after the payment has been reduced, the lender still has the option to foreclose. If collateral values are stable or increasing, the lender has nothing to lose. On the other hand, if collateral values are going down, the delay caused by a modification could mean that the lender will have lower recovery possibilities once the property is finally sold in foreclosure.
However, even with this “redefault” risk, it generally makes economic sense for a lender to try a modification first. If the lender is successful in rehabilitating the loan, the payoff will usually be much greater than any incremental losses in recoveries caused by several months of delay in the foreclosure sale. That is the reason that when banks hold loans in their own portfolios, they will aggressively work with distressed borrowers to restructure their obligations. A good example of this is the commercial real estate loans held in portfolios by community banks. The commercial real estate market has had its share of problems, but the fiasco being predicted by pundits in 2008 and 2009 has not materialized. One of the reasons is that so many of the loans are held in portfolio by community banks that have worked with their commercial borrowers to restructure loans, under guidelines issued by the regulators. In fact, if there has been a problem with commercial loan restructuring, it is that some banks have been too willing to accommodate borrowers, even when it is clear that the borrower’s finances are hopeless and the loan needs to be charged off. Contrast this to the spectacle we have seen in the residential housing market, where foreclosures, not modifications, have been the norm, and more than 6 million people have already lost their homes. Why the difference? A major culprit, again, is securitization.
What FDIC staff understood early on—frankly, before anyone else—was that the usual forces of economic self-interest would not result in the kind of wide-scale restructuring that was needed to avoid a massive wave of foreclosures. That was because through the securitization process, those who owned the mortgages were different from those who would be responsible for restructuring them and the legal contracts governing the modification process created economic incentives skewed in favor of foreclosure.
It has been in community banks’ economic self-interest to restructure their commercial real estate loans, because they own those loans. They have strong incentives to avoid incurring the steep losses associated with foreclosures. However, securitized loans are owned by a diverse group of investors with differing economic incentives. The investors do not service their own loans; that is, they do not collect the mortgage payments each month and take action against borrowers who become delinquent. Rather, they hire a financial institution to do the servicing. Loan servicing has become37 an increasingly concentrated business that is now performed principally by our largest banks or their affiliates.
Prior to the crisis, the job of a residential mortgage servicer consisted primarily of collecting mortgage payments and passing them on to investors. When a loan would occasionally default, the servicer would simply refer the loan to a foreclosure attorney. Servicers were not set up to deal with mortgage default because it happened so infrequently. Similarly, the agreements under which they operated compensated them based on a flat fee; they were not paid more for dealing with a delinquent loan, so their economic incentive was to do as little as possible with a troubled borrower. Indeed, during the go-go years leading up to the crisis, competition among servicers for the fees generated by the burgeoning securitization market intensified, driving fees down further and making the business one purely of volume, not of effective servicing. Not surprisingly, under that flat fee structure and in the face of intense competition, servicers never invested sufficient resources to deal with significant delinquencies. There are minimal costs associated with collecting mortgage payments from performing borrowers and passing them on to investors. However, when a loan becomes delinquent, working with a troubled borrower to restructure a loan can be a time-consuming, labor-intensive process, particularly if each modification is individually negotiated. Servicers were not compensated for making the extra effort, so why bother?
Actually, as we would soon discover, if anything, servicers had affirmative economic incentives to go to foreclosure quickly. That was because when a loan they serviced became delinquent, they were required to continue to advance the mortgage payments to the investors out of their own pockets. If they modified the loan instead of foreclosing, they would be reimbursed by the borrower slowly, over a period of years, by taking out a small part of the borrower’s new monthly payment. On the other hand, if they went to foreclosure, they were paid immediately, off the top, from foreclosure sale proceeds. If you were they, which would you do?
Why wouldn’t investors tell the servicers to modify loans? After all, if a foreclosure cost more money than a modification, it was the investors, not the servicers, who took the loss. But in point of fact, just the opposite happened, with some investors threatening to sue servicers over modifying loans. Why would investors want to sue servicers for trying to rehabilitate delinquent loans? After all, that would usually save them money over the cost of foreclosure. The answer to that question goes to the heart of what I believe was probably the single biggest impediment to getting the toxic loans restructured: the conflicting economic incentives of investors themselves.
Remember the tranches we discussed? As you will recall, most mortgage securitizations were set up to provide the senior tranche—the triple-A portion of the securitization—with substantial overcollateralization. What that meant was that if a mortgage defaulted, it had no impact whatsoever on the senior tranche—unless the defaulting mortgages exceeded 20 to 30 percent of the mortgage pool. However, here is the catch: because of the way in which many securitization documents were written, if, instead of a foreclosure sale, the loan was modified, the reduced mortgage payments flowed through to all investors in the securitization pool, meaning that everyone’s income was reduced, including that of the triple-A investors.
So again, what would you do if you were a triple-A investor? If a loan becomes delinquent and the servicer modifies it with
a 30 percent payment reduction, your portion of the payment flows from that mortgage will be reduced along with all the other bond holders. If, however, the servicer simply forecloses on the loan, even if the losses on foreclosure amount to 50 percent, you will still prefer the foreclosure because that entire loss will be absorbed by the lower tranches. From the standpoint of investors as a whole, it obviously makes more sense for the loan to be modified with a 30 percent loss instead of a 50 percent loss on foreclosure. However, from the standpoint of the triple-A bondholders, it makes more sense to foreclose. And the triple-A bondholders were more numerous and more powerful than investors holding the subordinate tranches.
Working with Jim Wigand, Mike Krimminger, and George Alexander, I decided that the best thing to do would be to get all stakeholders in a room together and try to hash out some type of agreement to start modifying subprime hybrid ARMs. Delinquencies on subprime hybrid ARMs were increasing quickly, and nearly half a trillion dollars’ worth of such loans were scheduled to reset in 2007 and 2008. The answer seemed obvious: eliminate the reset and simply extend the starter rate. In other words, convert the loan into a thirty-year fixed-rate mortgage, keeping the monthly payment the same as it had been during the starter period. We thought that investors—even Triple-A investors—should support such a step. We weren’t really proposing that their payments be reduced, just that they give up a payment increase that they had never had a realistic expectation of receiving. As previously discussed, hybrid ARMs were designed to force refinancings after two to three years, not to be paid at the higher rate for the life of the loan. Our data confirmed that the debt-to-income ratios on these loans were extremely high. Indeed, more than 90 percent of hybrid ARMs were refinanced at the end of the starter period. The number of borrowers who continued paying after reset was minuscule. Without some relief, subprime borrowers would default on a large scale, generating heavy losses for all bondholders, as well as the broader housing market.
