Financial Product Innovation

There are two main sorts of product innovations for the mortgage industry from the mid-1970s on. First was the proliferation of products for households that wanted to get mortgages but faced financial challenges in doing so.

Innovation was not straightforward. Financial institutions began to experi­ment with different kinds of new products. Some failed, and some succeeded. To be successful, financial institutions needed to make products to please two sorts of audiences. New financial products were built to attract buyers. House­holds that could not afford conventional mortgages were the main candidates for financial innovation in mortgages. The period of high inflation and slow economic growth in the 1970s and early 1980s was particularly challenging for households wanting to buy homes. The breakdown of the savings and loan mod­el of taking individual deposits in passbook accounts and using them to fund mortgages meant that new sources of capital needed to be found. Financial insti­tutions moved from tapping individual savers to trying to raise capital from the capital markets. This entailed creating products to sell to short-term and long­term investors.

Innovations in Mortgages

Many of the products required rewriting the rules that governed financial prod­ucts. Financial institutions who were interested in selling these new products had to get the government to agree to change the rules governing mortgages in order to build markets for these new products. Congress and the regulatory authorities consistently acted to promote changes in the rules governing finan­cial institutions to allow them to create new products. Here, I describe these products and briefly discuss the process by which they were brought to market.

The adjustable-rate mortgage (ARM) has a long history (Green and Wachter, 2005). Before 1930, at least half of American mortgages had interest rates that adjusted over time. But after the establishment of the conventional mortgage, it disappeared. During the high-inflation, high-interest-rate 1970s, state-char­tered banks began to experiment with mortgages that had adjustable interest rates. Attempts in 1971 and 1974 by the Federal Home Loan Bank Board (FHLBB) to authorize residential ARMs met with stiff resistance by Congress (Cassidy, 1984). Opposition was widespread among consumer groups and labor unions that feared borrowers would be subjected to unmanageable increases in their mortgage payments.

By the end of the 1970s, however, as the condition of the thrift industry rap­idly deteriorated, the political climate began to change (Barth, 2004). In Decem­ber 1978, the FHLBB allowed federal savings and loan institutions in California to originate variable-rate loans in competition with state-chartered institutions.

The ARM solved several problems. For banks that wanted to hold on to mort­gages, it meant that mortgagors would assume the risk of rapidly changing in­terest rates. If rates went up, banks were able to cover their cost of capital. The buyers of ARMs during the late 1970s and 1980s were able to get mortgages that otherwise would have been unavailable. In many cases, the ARM came at lower rates than fixed-rate mortgages precisely because they could move with inter­est rates. Many of these mortgages came with initially low rates (called “teaser rates”) whereby the loan would start out very low and adjust more rapidly. From 1980 on, the use of ARMs was very much related to overall interest rates (Mo- ench et al., 2010). When rates fell, households would buy fixed-rate mortgages. Many who had ARMs would refinance when interest rates fell as well. As interest rates rose in the early 1980s, the share of total mortgages that were adjustable went from 29 percent in 1981 to 70.1 percent in 1985 (Peek, 1990). The use of ARM products has risen and fallen subsequently in response to interest rate changes (Green and Wachter, 2005).

Several features of ARMs are negotiated between financial institutions and mortgagors. Of course, one of the most important is the beginning interest rate. Equally significant is the length of time that the ARM is scheduled to remain unchanged and how much it can be reset. The reset rate is usually linked to some kind of market index such as the one- or three-year Treasury bill or what used to be the London Interbank Offer Rate (LIBOR) index.

The 1980s and 1990s produced other innovations that further expanded the range of mortgage choices in order to expand the number of households that could afford to buy houses. Financial institutions changed the level of down pay­ment necessary, the credit requirements for loans, and the amount of money that households could borrow. These changes reflected the relatively high inter­est rates in place over this period as well as the increasing prices of houses and the desire of financial institutions, particularly commercial banks, to expand the size of the market for mortgages. While consumers benefited in having more choices and opportunities to get a mortgage, financial institutions found this type of expansion of the market quite lucrative. Anyone who took out one of the nonconventional mortgages could expect to pay a higher interest rate that was justified by arguing that such loans were riskier for financial institutions. Mort­gagors also had to pay additional fees to get the loan and, in some cases, take out mortgage insurance to guarantee the loan. Financial institutions eventually dis­covered that giving credit to people who had had some impaired credit history or special financial circumstances turned into a lucrative business.

I will use the term nonconventional to describe all these types of loans that relax the terms of the conventional mortgage. Alt-A mortgages refer to mortgag­es where a household has some impaired credit history or lack of verifiable in­come.

Two other forms of nonconventional mortgages were also pioneered in the 1980s and 1990s that catered to households with special circumstances. Jumbo mortgages are loans for households that have high credit quality but need a loan amount above what is called the “conventional conforming loan limits.” These are standards set by the two government-sponsored enterprises, Fannie Mae and Freddie Mac, that refer to the maximum value of any individual mortgage they will purchase from a lender. Jumbo mortgages were initially marketed to high-income households who wanted to buy luxury homes. Such households could afford higher monthly payments but may not have had sufficient down payments to bring down the loan amount to a conventional level. As high prices inflated in the 1990s and 2000s, particularly in urban areas on the coasts, many middle-class households found themselves priced out of conventional mortgag­es. To get a loan, they needed to have a jumbo mortgage.

Home equity loans (HELs) are loans that borrow against the equity value of one's home. Before the 1980s, if a household had financial problems, they might have taken out a second mortgage. The companies that pioneered these loans usually marketed them to lower-income households who were in a more pre­carious financial situation. This would allow them to stay in their homes while waiting for their financial circumstances to improve. But during the early 1980s, financial institutions realized that households had accumulated equity in their homes. By offering people loans to extract that equity, financial institutions could create a financial instrument that was backed by collateral—that is, the value of the house. Several marketing campaigns used the term home equity loan instead of second mortgage, and it stuck. These loans also tended to have higher interest rates attached to them even though they were backed by the value of the home.

Rising home values in the 1980s increased the amount of homeowner equity in the United States. This asset became an attractive source of financing after the passage of the Tax Reform Act of 1986 (TRA). TRA prohibited taxpayers from deducting interest on consumer loans, such as credit cards and auto loans, while it allowed them to deduct interest paid on mortgage loans secured by their principal residence and one additional home. This change provided an incentive for homeowners to take home equity loans and use the proceeds to pay off con­sumer debt, finance the purchase of cars, or pay for their children's education. Another bonus was that all of these other forms of borrowing typically come with higher interest rates than home equity debt.

All of these qualifications to the conventional mortgage were possible only in a more deregulated financial services market. Two laws enacted in the ear­ly 1980s, the Depository Institutions Deregulation and Monetary Control Act (DIDMCA) and the Alternative Mortgage Transaction Parity Act (AMTPA), made it possible for lenders to originate mortgages with prices and features previously prohibited by individual states (Temkin et al., 2002). In particular, they made it possible for lenders to originate loans with higher interest rates and a broader range of terms than previously allowed under the various state lending laws.

DIDMCA, adopted in 1980, helped set the stage for the growth of subprime lending. The law produced sweeping changes in the way that banks operated across the country. It forced all banks to abide by the federal government's rules. It also allowed for cross-border mergers of all kinds. It also removed any ceil­ings or floors on interest rates that banks could charge or offer to customers. It increased deposit insurance for all banks and credit unions from $40,000 per account to $100,000. By deregulating loan rates through the preemption of state interest rate caps for first lien loans on a borrower's house, the law allowed banks to create whatever mortgage products they wished. DIDMCA was an important element in the subprime market's growth because it allowed lenders that orig­inate de facto second mortgages and consumer loans to make more expensive first lien loans, the most common type of subprime loans.

AMTPA was adopted in 1982 and was part of the Garn-St. Germain Act. It preempted state laws that restrict a number of alternative mortgage features that are important subprime lending elements. Its intent was to increase the volume of loan products that reduced the up-front costs to borrowers in order to make homeownership more affordable. Specifically, the legislation allows lenders to originate mortgages with features such as variable interest rates, balloon pay­ments, and interest-only mortgages in any “loan or credit sale secured by an interest in residential real property made, purchased or enforced by covered lenders” (National Mortgage News, 1982).

The wave of innovation in the production of types of mortgages was driven first and most importantly by the difficult economic times of the 1970s and early 1980s. High inflation and high interest rates made it imperative for the savings and loan industry to find new ways to fund mortgage origination and new strat­egies to charge more for mortgages but still be able to do originations. They were facilitated by a series of pieces of legislation that were friendly toward the industry but also more broadly toward commercial banks and mortgage brokers and banks who were interested in growing their share of mortgage originations.

While this new legislation did not save the savings and loan business model in the end, it did provide the legal underpinnings for whole new kinds of mortgages focused on segmenting the housing market. The 1980s and 1990s were an era for experimentation in mortgage finance for originations. As one market peaked or dried up, innovators looked for new markets to whom they could sell mortgage products. For example, jumbo mortgages were originally marketed to high-in­come households. But as the price of housing increased during the 1990s and early 2000s, financial institutions began to market this product, which earned higher interest, to more middle-class households, thus expanding their markets. The innovators in all of this were not the savings and loan banks by and large but a new group of mortgage banks such as Countrywide Financial and commercial banks such as Bank of America and Citibank.

The second problem presented to the savings and loan industry during the 1970s was the crisis of obtaining funds to originate mortgages. Their model of tak­ing deposits from retail customers failed during the 1970s to provide savings and loans with enough funds to engage in creating mortgages and holding the mortgages. Regulation Q was being undermined by financial innovators who created money market funds, certificates of deposits, bond mutual funds, and other forms of investment so that ordinary people could make investments to take advantage of high interest rates. The savings and loan industry lobbied and got Regulation Q repealed in the Garn-St. Germain Act of 1982. They also man­aged to have depository insurance raised. This allowed them to access capital at whatever level of interest they needed in order to attract funds. This, of course, turned out to be a disaster, as in order to pay out higher interest rates, they had to engage in riskier ventures (Barth, 2004).

The savings and loan banks were not the only financial institutions to bene­fit from financial deregulation. The GSE model of using the capital markets to raise funds for mortgages by creating MBSs took over a decade to come into being. Many of the problems of creating MBSs ultimately required the govern­ment to create new rules to allow both sellers and buyers of MBSs to participate in the new market. In the wake of the collapse of the savings and loan banks, the GSEs took over the central role of organizing the mortgage market. This takeover meant that from the 1980s on, money for mortgage origination was increasingly being raised in the capital markets. It was not just the GSEs who went to these markets but also many financial institutions that were also originating and secu­ritizing mortgages. During the 1980s, many financial institutions who originated mortgages turned around and sold them either to wholesalers or directly to the GSEs to have them packaged into MBSs. Many of these financial institutions had lines of credit from the capital markets that they would use to do origination.

While some of this capital came from borrowing from banks or relying on the firm's own capital, some of it also came from the commercial paper market where money could be borrowed short term (Post, 1992; Anderson and Gascon, 2009). The commercial paper market grew from $125 billion in 1980 to almost $500 bil­lion in 1990 and a bit over $1.5 trillion in 2000. Most of this increase came with the introduction of the asset-backed commercial paper market in 1983. Before that, most of the activity in the market was for nonfinancial corporations to borrow money short term to fund inventories and meet payrolls. But banks increasingly used the market to borrow money short term to fund their financial activities.

The main products that came out of mortgage originations were mort­gage-backed securities (MBSs), collateralized debt obligations (CDOs), and credit default swaps (CDSs). Each of these products was created to interest in­vestors in owning bonds based on mortgages. MBSs and CDOs are distinct but overlapping categories. Mortgage-backed securities are any kind of asset-backed security where the underlying assets are mortgages. They may have one class (tranche), as in the case of pass-through securities, or many classes (tranches). CDOs can have underlying assets of any kind of debt (bonds, mortgages, even other securities). They always have multiple tranches with different priorities of payments. An MBS with a CDO-like structure is called a CMO (collateralized mortgage obligation) or sometimes an MBS-CDO. When a CDO contained a myriad of assets, it was frequently referred to as an ABS-CDO, where ABS was an abbreviation for asset-backed security.

The word tranche is French for slice, section, series, or portion and is a cog­nate to the English word trench (Fender and Mitchell, 2005). Bonds are divided into tranches in order to separate out how risky various parts of the underlying assets might be. The tranches are identified as those that are the most senior to most subordinate. The more senior-rated tranches have higher bond credit ratings than the lower-rated tranches and are thus considered safer than those below them. For example, senior tranches may be rated AAA, AA, or A, while mezzanine tranches might be rated BBB or below. Tranches with a first lien on the assets of the asset pool are referred to as senior tranches and are generally saf­er investments. Typical investors of these types of securities tend to be the GSEs, insurance companies, pension funds, and other risk-averse investors (Fender and Mitchell, 2005).

One of the most perplexing aspects of these distinctions is that CDOs could contain a mix of assets. CDOs originated to securitize assets from a variety of asset groups (Tett, 2009; MacKenzie, 2011). A particular CDO might include mortgages, credit card debt, and funding for planes rented by leasing companies. Since ABS-CDOs could contain any kind of financial asset, financial institutions began to package together tranches of other CDOs that they could not otherwise sell, either because the tranches were too risky or because they were too safe and paid too low a return, what were called “super senior tranches” (MacKenzie, 2011). When CDOs were exclusively made up of other CDO tranches, they were referred to as CDO-CDOs or CDO-squared. They could also sometimes, con­fusingly, be called ABS-CDOs.

In the case of MBS-CDOs, after 2001 financial institutions found that bond­holders preferred more highly rated tranches over lower-rated tranches. This left them with lower-rated tranches that they could not sell. The CDO process al­lowed financial institutions to rebundle these unsalable BBB-rated mezzanine MBS-CDOs into highly rated securities. They did so by producing new tranches

that allowed purchasers to buy whatever level of risk they chose (MacKenzie, 2011). How could riskier tranches end up being rated AAA? Given that not all of the mortgages from MBS-CDOs that were rated BBB or below in a CDO-CDO would fail (indeed, only a minority ever did), buyers could pay to buy a level of risk. This process could leave some tranches unsold. These could then be repack­aged again into CDOs (sometimes referred to as CDO-cubed). While pricing a CDO-CDO can be difficult due to the disparate income streams from which it is constituted, for those based on mortgages, at root it is a claim on MBS-CDO payments, which in turn are claims on income from mortgage payments.

A credit default swap (CDS) is an agreement that the seller of the CDS will compensate the buyer of some asset in the event of a loan default (by the debt­or). In essence, the seller of the CDS insures the buyer against a loan defaulting. The buyer of the CDS makes a series of payments to the seller and, in exchange, receives a payoff if the loan defaults. Early trades of the CDS were pioneered by Bankers Trust in 1991. The modern standard form of the CDS was created by Blythe Masters, who worked at J.P. Morgan in 1994 (Tett, 2009). It has been extensively used by financial institutions and corporations to modify all kinds of risks. One of the interesting features of a CDS is that one does not have to own a particular asset in order to create a CDS with some other party. For example, one could take out a CDS agreement about a particular CDO even if one did not own it.

In the case of MBS-CDOs, buyers of CDSs in the beginning used CDSs to protect themselves from potential downturns in the housing markets. By buying CDSs, holders of MBS-CDOs not only lowered their risks, but regulators would then allow them to hold less capital for holding those MBS-CDOs. If a finan­cial entity bought AAA-rated CDOs that they then took out a CDS against the risk it would default, they had to hold little or no capital for that purchase. This allowed financial institutions to increase their purchases of CDOs and have a higher level of leverage. CDSs were also used in 2006-2008 by certain investors at hedge funds to bet against the CDO market. By buying CDSs for CDOs that investors thought were likely to fail, investors could profit if and when they did (Lewis, 2010).

The history of the creation of MBS-CDOs goes back to the 1960s (Quinn, 2019). The modern securitization market was a direct invention of the federal government. Their goal was to increase homeownership by increasing the pool of money available for investment in mortgages (Sellon and VanNahmen, 1988). In particular, the Johnson administration was concerned about whether the savings and loan model of mortgages would be able to provide enough money to fund mortgages for the baby boom generation that was coming of age in the 1960s. One solution that was discussed was to expand the FHA and allow the government to be the lender to millions of Americans. The problem was that the government did not want to get into the business of being the lender of last resort. It would mean that the government would absorb billions of dollars of debt into a budget already strained by the Vietnam War (Ranieri, 1996).

To deal with this, the Interagency Committee on Housing Finance was con­vened in 1966, and it was followed by the Mortgage Finance Task Force in 1967. The latter, led by Federal Reserve governor Sherman Maisel and James Dusen- berry of the Council of Economic Advisers, decided that the preferred solution would be to bolster the secondary market for mortgages. This would allow the federal government to sell off the mortgages on its own books and help other lenders do the same (Sellon and VanNahmen, 1988; Green and Wachter, 2005). The goal was to free up capital held in reserves by private companies, generate new capital from sales to increasingly important investors such as mutual and pension funds, and thereby create efficiencies, savings, and new sources of capi­tal that could be passed on to home buyers (Sellon and VanNahmen, 1988).

The federal government undertook sweeping housing regulation reform from 1968 to 1970, which began with the privatization of Fannie Mae and two years later resulted in the creation of the first modern securitization instruments. Fan­nie Mae had long been in the business of providing stability for the housing market by promoting the secondary market. Specifically, it purchased mortgages during market downturns in order to encourage renewed lending. However, go­ing forward, it would do so as a private entity.

To make sure that the newly private agency was successful, Ginnie Mae was created in 1968 as a new governmental agency that would officially support the market on behalf of the federal government, mostly by providing guarantees for governmental debt held by and sold through Fannie Mae. Even as the govern­ment took steps away from direct ownership of mortgages, it found it needed to maintain a directive presence in order to secure investor confidence, and the step toward privatization of this governmental agency was offset by a complex set of governmental protections. Two years later (in 1970), Freddie Mac was created as a private entity modeled after Fannie Mae. Given a similar charter and privileg­es, it was to compete with Fannie Mae and help create a secondary market for mortgages originated by thrifts.

As government-sponsored enterprises (GSEs), Fannie Mae and Freddie Mac follow a congressional charter that requires them to support the secondary mar­ket for housing by promoting liquidity and stability. Their debt and obligations are not backed by the US government, but they have special privileges and tax exemptions and “conditional access” to a $2.5 billion line of credit from the gov­ernment, and they are directly regulated by a special office in HUD. In 2003, the Securities and Exchange Commission (SEC) received authority over them after a set of scandals. Moreover, that their debt commonly trades at the level of govern­ment bonds suggests that investors believe that they are simply too big and im­portant for the US government to ever let fail. In this way they carry an implicit, though not legal, guarantee by the US government. The hybrid public-private nature of the GSEs is best thought of as the place where the US government most directly intervened into structuring the housing market.

It was within this configuration of public and semiprivate agencies that the securitization market was born. They started with the idea of auctioning off Fan­nie Mae's mortgages and spent a good year figuring out how to make such a plan work. They decided in the process that some sort of government guarantee was crucial in order to secure investor confidence. Their efforts came to fruition when the first mortgage-backed security was issued in 1970. A group of mort­gages backed by the FHA and VA were given a government guarantee of timely payment by Ginnie Mae and then sold off as bonds on the capital markets (Sellon and VanNahmen, 1988).

These deals were called pass-throughs, as the ownership of the mortgages was literally transferred to the investors along with their payments. Investors were mainly the thrifts, private banks, and insurance companies. They had been enticed with a chance to take a position in the housing market without taking any credit risk. There were, however, significant limitations to the use of pass­throughs. Throughout the 1970s, officials at Ginnie Mae and executives at Fannie Mae and Freddie Mac worked closely with investment bankers to develop new financial instruments that solved these problems.

The private sector was a partner in these efforts. Salomon Brothers was the leading investment bank interested in promoting MBSs. Robert Dall, a trader at Salomon Brothers, helped create the first private MBS with Bank of America in 1977. The deal was for $100 million. It was the first MBS that experimented with something like tranching. In 1983, people from both industry and government created the first securitization structures with tranches as they are now used, creating the collateralized mortgage obligations (CMO).¹ These multiclass issues were incredibly flexible and incredibly successful. Michael Lewis writes, “The CMO burst the damn between several billion investable dollars looking for a home and nearly two trillion dollars of home mortgages looking for an investor” (1990: 136).

Over the following decades, deals would be parsed in increasingly complicat­ed ways; many would have over forty different classes of debt, and at least one contained more than ninety tranches (Kendall, 1996; Kochen, 1996: 109). Soon, new kinds of investors were entering the market (Sellon and VanNahmen, 1988), reportedly attracted by instruments designed with their preferences in mind (Ranieri, 1996; Fink, 1996). Within twenty-five years, the financial structures proposed by the government officials in the Mortgage Finance Task Force, and initially bolstered with government guarantees through Ginnie Mae, grew into a multitrillion-dollar trade of sometimes staggeringly complex instruments.