read case and answer four questions
3-4 pages (no more than 4 pages)
Questions are in the end.
Savings and Loans and the Mortgage Markets
This note provides a brief overview of the history of the Savings and Loans, the Savings and Loans crisis of the 1980s and early 1990s, and the creation of the mortgage markets in the U.S. It also explains briefly the most common types of mortgage-backed securities (MBS) available.
Savings and Loans Before the Crisis
Savings & Loans (S&Ls) first appeared as financial intermediaries in the early to mid-1800s. The principal role of the S&Ls was to take in retail deposits (providing a riskless, liquid, short-term savings vehicle for large numbers of small savers) and originate long-term fixed-rate mortgage financing for residential homeowners. S&Ls were originally voluntary associations. People pooled their money to make it possible for members of the association to get mortgages on their homes without having to resort to loan sharks or to other forms of usury. The first documented mortgage transaction occurred in 1831, when Isaac Shallcross, as secretary of the first S&L made a $381 mortgage loan to lamp-lighter Comly Rich. Rich bought a house that still stands in the Frankford suburb of Philadelphia.
S&Ls were (and still are) community-based, and local in nature. This is partly due to tradition and to prohibitions against out-of-state lending, but it is also because mortgage origination is mostly a local business. People with knowledge of local real-estate market conditions need to participate in the mortgage application and appraisal process, as well as deal with the loan and property in case of default. Thus, even though the source of funds might be distant, agents of the lender need to be locally based, so they will have access to the necessary data to make an informed decision. There is evidence that information and incentive problems increase as the distance between the lender and the borrower grows. One 1948 study of savings banks, for example, found a direct correlation between loan loss rates and the distance of the loan from the home office.
Another important aspect of the thrifts before the S&L crisis was their capital structure. The voluntary nature of the early thrifts gave rise to an unusual corporate form. The vast majority of thrifts were not established as stock corporations, but as mutual associations. In a mutual association, the thrift is organized as a cooperative owned by its members. Member’s deposits represent shares, and shareholders vote on company matters and receive interest on their deposits in the form of dividends. Mutual associations are not permitted to issue stock, and deposits (i.e. “shares”) are technically subject to a waiting period before withdrawal. In practice, withdrawals are usually on demand, with the thrift’s liquidity assured by its ability to borrow from the Federal Home Loan Banks (described below) using home mortgages as collateral. Prior to the S&L crisis, regulations constrained most thrifts to be mutually owned.
The Rise of the Mortgage Markets
Attempts in the 1800s to overcome the community-based lending structure of S&Ls failed. It wasn’t until the Great Depression in the 1930s, that permanent change was forthcoming. Seeking the revival of the housing industry, Herbert Hoover’s administration formed the Federal Home Loan Bank System (FHLB), a dozen district public-purpose “banks”, owned (like the Federal Reserve Banks) by their member institutions. The FHLB banks were controlled by a national Board (the FHLBB) appointed by the President and confirmed by the Senate. The Housing Act of 1934 added a Home Owners Loan Corporation (later renamed the Federal National Mortgage Association, or Fannie Mae), which bought mortgages from S&Ls. The Act also empowered the FHLBB to issue national charters for S&Ls (which despite their national charter were restricted to their state of origin)1 and established the Federal Housing Administration (FHA).
It was through the FHA that the first steps in the formation of national mortgage markets were taken. The FHA offered credit insurance on mortgages that met the underwriting standards it established. It also developed the long-term self-amortizing mortgage that became the standard structure in residential financing. The FHA not only provided means for reducing credit risk to lenders, but also set the stage for the standardization of the lending criteria and mortgage terms, both of which were fundamental in the later development of the securitization process.
While these were important advances, they did not expand the mortgage markets from a local to an inter-regional or long-distance market. Until the mid-60s, the S&L business was secure, profitable, and dull. S&L managers operated under the well known 3-6-3 rule: borrow from depositors at 3%, lend to home buyers at 6%, and arrive at the golf course by 3:00 PM. In 1970, savings institutions accounted for 60% of the outstanding home mortgages. The large majority of these institutions made mortgage loans only within their home state, and often within a more narrowly defined home region or lending area. The volume of mortgage loans available within a particular region was limited by the funds available to the local depository institutions.
By the late 1960s, when interest rates began to rise and became more volatile, there were 4,508 S&Ls, with total assets of $124.6 billion. This rise and increased volatility of interest rates caused difficulties for the S&Ls, which make their profit from interest rate spreads — the difference between what they collected in mortgage-loan interest and what they paid to depositors. If interest rates drop, borrowers have the option of refinancing, so the thrift gets its money back sooner than it wants and has to reinvest at a lower rate. Of course, the thrift also pays lower interest rates on the deposits, offsetting some of this interest rate risk. However, as interest rates rose, borrowers held on to their (fixed-rate) mortgages, and the thrifts were stuck paying more for deposits than they earned on their loans.
Compounding the problem, interest rate ceilings, such as the Federal Reserve’s Regulation Q, prevented S&Ls from paying competitive interest rates on deposits. Thus, every time interest rates rose, substantial amounts of funds were withdrawn by consumers for placement in instruments with higher rates of return, such as mutual funds. This process of deposit withdrawal (“disintermediation”) and the subsequent deposit influx when rates rose (“reintermediation”) left S&Ls highly vulnerable. S&Ls were additionally restricted by not being allowed to enter into businesses other than accepting deposits and granting home mortgage loans.
1 Even more importantly, they were only permitted to lend within a fifty-mile radius of a branch office.
In response to the resulting increase in lending institutions’ cost of funds and the growing disintermediation problem, the Federal National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage Association (Freddie Mac, formed in early 1970 by the FHLBB) were authorized in 1970 to purchase pools of mortgages and to issue securities collateralized by the pools. Fannie Mae and Freddie Mac, as federally-sponsored entities, guaranteed the payment of interest and principal. The Government National Mortgage Association (Ginnie Mae, created by the Johnson administration in the 1960s) also supported securitization by insuring private securitization of FHA- guaranteed mortgages. These securitized pools of mortgages are known as Mortgage Backed Securities (MBS) pass-throughs. This way, thrifts could package and sell their mortgage loans, removing them from the balance sheet and injecting liquidity to the institution.
Until 1977, all pass-through securities carried the full backing of the US government. In 1977, the first private mortgage pass-through not guaranteed by the government was issued by Bank of America. Nowadays, commercial banks, investment banks, and others package privately-issued mortgages and issue securities backed by pools of mortgages. There are a number of credit enhancement techniques: the issuer of the securities can guarantee the MBS if it has a high enough credit rating, or credit insurance can be purchased from a specialized insurer.
The issuance of these mortgage securities made investing in the mortgage market more appealing to an even greater institutional investor base, which had found the investment in individual mortgages a risky proposition. As issuers of these simple MBS pass-throughs became more comfortable with the securitization process, other types of mortgages, such as adjustable rate mortgages, were also securitized.
Mortgage securitization has transformed a highly localized lending market into one in which country-wide and even international lending is common. Of the $1 trillion of mortgages outstanding in the United States in 1993, 63% of them were securitized. By 1996, the amount outstanding was $3.8 trillion (which corresponded to 32% of all public and private debt outstanding), of which over $3 trillion was securitized.
Mortgage securitization has made the transfer of resources much more efficient by eliminating the link between the origination (which is still a local business) and the funding of the loan. Additionally, securitization has also facilitated easier matching of cash flow requirements over time. The typical long-term fixed rate mortgage loan offers is advantageous to the borrower. However, this single instrument might not be the optimal investment vehicle for investors, or even for the savings institutions that originated the mortgages in the first place. By splitting the securitized mortgage pools into instruments of different maturities,2 it has been possible to design different mortgage instruments that appeal to different investors.
For example, pension funds have a need for very long-duration assets to match their very long liabilities. With MBS, it is possible to create very long-duration assets by separating the longest- dated cash flows from the mortgage pool. In a similar fashion, instruments were created to meet the needs of commercial banks, pension funds, and other financial intermediaries. As a measure of the relevance of MBS to other financial intermediaries, in 1993, only 11% of MBS were owned by savings and loans, while 26% were owned by banks, and 13% by pension funds.
The Savings and Loans Crisis
Also in response to the problem of disintermediation in the late 60s, state and federal governments began the deregulation of the industry and the application of measures to ensure the viability of the S&Ls.
For example, after a number of California thrifts failed in 1966 after the Fed raised interest rates to control Vietnam war-induced inflation, the FHLB, in response, allowed S&Ls to pay a 0.25% differential on insured deposits over the maximum rates permitted to the banks under Regulation Q. Another important step was the gradual lifting of restrictions (beginning in 1970 in California) on the issuance of adjustable-rate (ARM) mortgages.
Regulatory “forbearance” was also begun in this period with respect to loosening restrictions on S&L assets. For example, the State of Texas allowed property development loans of up to 50% of a bank’s equity in 1967, and in 1978, the Financial Institutions Regulatory and Interest Rate Control Act allowed S&Ls to invest up to 5% of assets in each of land development, construction, and education loans.
In 1979, the problems facing the S&L industry intensified. Oil prices doubled, inflation moved into double digits for the second time in five years, and the Federal Reserve decided to target the money supply in order to control inflation, letting interest rates not only rise, but become even more volatile. As the prime rate hit 21%, certificates of deposit issued by the S&Ls went for as high as 15%, while their income came in the form of long-term lending of fixed rate mortgages, many made at 7% or 8%. Additionally, Regulation Q led to a large drain of deposits out of thrifts and into unregulated investment vehicles (such as Eurodollar deposits and mutual funds), further exacerbating their problems.
In response, the government enacted statutory and regulatory changes that gave the S&L industry new powers so they could enter new areas of business and return to profitability. For the first time, the government approved measures intended to increase S&L profits as opposed to promoting home-ownership. In March 1980, the Depository Institutions Deregulation and Monetary Control Act (DIDMCA) was enacted.
DIDMCA was an initiative aimed at eliminating many of the distinctions among different types of depository institutions and ultimately removing interest rate ceilings on deposit accounts. DIDMCA authorized all depository institutions to offer interest-bearing checkable deposits, such as NOW accounts, and mandated the elimination of Regulation Q by 1986. DIDMCA also allowed the Federal Savings and Loans Insurance Corporation (FSLIC), the government-sponsored S&L deposit insurance corporation, to raise insurance deposit limits from $40,000 per account to $100,000. This increased the protection of the thrifts by the government, but had the added perverse effect of allowing thrift managers to “gamble” with insured deposits.3 Finally, DIDMCA allowed insured thrifts to invest in broader, riskier classes of assets, such as real estate and ADC (acquisition, development, and construction) loans. Removing these asset restrictions was viewed by some as a step towards deregulation, which would allow thrifts to diversify and grow their way back to health, but was seen by others as a reduction in the intensity of monitoring of insured thrifts, which would cause further problems down the road.
3 This is an example of the well-known “moral-hazard” problem. In this case, the government socialized and distributed the risk of questionable loans, while leaving the potential rewards private. Another illustrative example is the increased incidence of speeding or reckless driving as cars with enhanced safety measures such as anti-lock brakes or airbags become more common.
This regulatory forbearance, begun with DIDMCA, and aimed at having S&Ls grow themselves into profitability, continued through the early 1980s. The regulations and statutes enacted were many and varied, but five of them stand out as being particularly relevant.
1. In November 1980, the Federal Home Loan Bank Board (FHLBB) reduced the net worth requirement for FSLIC-insured S&Ls from 5 to 4% of total deposits, and removed the limits on the amount of brokered deposits4 an S&L could hold. In January, 1982, the FHLBB further reduced the net worth requirement from 4 to 3% of total deposits. Additionally, S&Ls were allowed to meet the low net worth standard not in terms of generally accepted accounting principles (GAAP) but of the more liberal regulatory accounting principles (RAP). This relaxation in accounting standards and net worth requirement allowed some of the insolvent thrifts to become statutorily (but not economically) solvent.
2. The approval of “Memorandum R-49”, by the FHLBB in September, 1981. The memorandum encouraged “loss deferral.” Loss deferral allowed S&Ls to sell loans on which they showed a loss and amortize the loss over the life of the loan. An additional “safe harbor” clause was inserted to allow the S&Ls to show the total amount of the loss only for tax purposes when the assets were sold (thereby generating a tax credit for the year.) By encouraging S&Ls to dispose of bad loans and mortgages, funds could be released for ostensibly safer and more profitable investments. However, the new loans issued were not necessarily safer. Under the new rules, S&Ls could buy even riskier investments, such as below-investment grade bonds (“junk bonds.”)
3. The Garn-St. Germain Depository Institutions Act of 1982 further expanded the classes of assets that a thrift could invest in, and changed the procedures for rescuing insolvent S&Ls through merger. The Act changed the FHLBB accounting rules to permit an S&L that acquired another to capitalize as goodwill the difference between the net worth of the acquired institution and the price paid for it, and to allow the amortization of goodwill to occur over forty years. The result was that insolvent, money-losing S&Ls would be worth more to an acquirer than sound ones, because the goodwill that could be booked in such a transaction would permit a larger increase in the acquirer’s statutory capital, allowing greater opportunities of expansion of the loan portfolio.
4. The Garn-St. Germain Act also included a provision that legalized and encouraged mutual associations to convert voluntarily to stock ownership. The proponents of stock conversion believed that the Act represented a chance to bolster the thrift’s capital positions (through equity issuance) and would lead to higher efficiency through better monitoring of managers by stockholders. Before the passage of the Act, 77% of thrifts were mutual associations, and 23% were stock owned. Stock owned institutions controlled about 30% of the industry’s assets. Between 1982 and 1986, close to 500 thrifts converted to stock ownership, raising over $5.3 billion in equity capital. Also, 75% of new institutions opted for stock ownership. By 1988, over 70% of the industry’s assets were held in stock owned institutions. Consistent with the theory of equity holder’s incentives,5 stock-owned thrifts grew faster, had higher leverage, went out of the local market to obtain brokered deposits, and sold more of their mortgages for securitization than did mutuals. Stock-owned thrifts then used the money raised from equity issues, mortgage sales, and brokered deposits to fund riskier portfolios than those of mutuals.
5. The FHLBB ruling in November 1982 that allowed “appraised equity capital.” An S&L that owned commercial property could credit into capital any increase in the appraised value of that property. This led to a perverse incentive to over- appraise properties in thrift’s balance sheets in order to increase its statutory capital.
4 Brokered deposits are funds packaged by brokerage houses in units of $100,000, to be eligible for deposit insurance. They are then placed in thrifts and commercial banks as deposits, earning market rates of interest, and serving as funding sources for the institutions.
5 In a stock-owned thrift, the shareholders are residual claimants, and their claims can be transferred via a sale of stock. Specifically, shareholders hold the equivalent of a call option on risky assets. As a result, they have incentives to increase the riskiness of the underlying assets and to take on additional debt. The shareholders have a strong incentive to maximize the value of the residual claim at the expense of debtholders or a guarantor (e.g. the FSLIC.) In contrast, in a mutual association, the depositors are not a distinct class. There is no transfer of shares, and only the face value of the deposit can be redeemed when an account is closed. Depositors have weak incentives to increase the riskiness of the assets or to take on additional debt.
Much of the money raised by thrifts was invested in commercial and residential rental real estate ventures. Although these ventures might have been profitable by themselves, the tremendous increase of these transactions by thrifts, coupled with the use of insured deposits to fund these ventures, led to unjustified risk-taking. Even though regulatory forbearance was reversed in the late 1980s, enough S&Ls had failed through disintermediation, excessive risk-taking, compounded losses, and fraud, that by January 1987 the General Accounting Office (GAO) declared the FSLIC (which had to rescue the depositors of the failed thrifts) insolvent by at least $3.8 billion. By 1989, when the Financial Institutions Reform Recovery and Enforcement Act (FIRREA) was passed, the number of FSLIC-insured S&Ls had fallen from 4,002 in 1980 (with total assets of $618.5 billion) to 2,930.
FIRREA abolished the FHLBB and FSLIC, and switched S&L regulation to the newly created Office of Thrift Supervision. Deposit insurance functions were shifted to the FDIC, and a new entity, the RTC, was created to liquidate the insolvent S&Ls. The RTC, which closed its doors in December 1995, eventually closed or merged 747 thrifts, protected 25 million depositor accounts worth $220.6 billion, and sold off more than $465 billion in assets, including 120,000 real estate properties. The final cost of the bailout was estimated at $145 billion, with the RTC’s share totaling about $90 billion.6 By 1995, only 1,478 S&Ls remained, with total assets of $777.4 billion.
The Future of Savings and Loans
Even while the government was, at great expense, trying to save the thrifts in the early 1990s, both of the thrifts’ economic functions of taking deposits and issuing mortgages were being taken over by other institutional mechanisms. The creation of MBS, issued initially to help the thrifts, led to the creation of a national mortgage market that then allowed mutual funds and pension funds to become major funding alternatives to the thrifts. These markets also allowed the entry of investment banks and mortgage brokers to compete with the thrifts for the origination and servicing of loans and mortgages. On the deposit side, mutual funds and other investment vehicles competed directly with thrifts and commercial banks for deposits, further eroding the thrifts’ ability to generate new loans.
Mortgage securitization has had a tremendous impact on the structure of the S&L industry. The largest originator of mortgages in 1993 was not a thrift or a financial intermediary, but Countrywide Credit Industries, Inc., a mortgage broker. Savings institutions still originated and held mortgages for homes in their communities, but by 1993 they held only 15% of the mortgages outstanding, down from 60% in 1970.
There were other issues facing the thrift industry in the mid 1990s. The S&L crisis weakened the industry. The healthy and well-managed thrifts that remained have been left to service $8 billion in bonds issued in 1987 as part of the initial recapitalization of the FSLIC. This alone will cost $780 million a year for the next 24 years. Also, the deposit insurance fund created to replace the FSLIC, the Savings Association Insurance Fund (SAIF) had to be recapitalized. This cost was estimated at around $14 billion in 1996 dollars, equivalent to a quarter of the equity base of the country’s remaining 1,500 S&Ls. The recapitalization of the SAIF was expected to take until the year 2003.
This additional burden on the S&Ls could be big enough to shift deposits from S&Ls to banks. With lower insurance premiums, banks were able to offer higher deposit rates. Even after the SAIF is recapitalized, the servicing of the long-term bonds issued in 1987 will continue to be a burden on the remaining S&Ls.
Even if a solution is found to the costs of insurance and of servicing the bonds, the future for thrifts is unclear. One problem is competitiveness. As an industry, S&Ls made a return on assets in 1995 of 56 basis points, less than half the 115 basis points of commercial banks. That reflected in part the lower risks (and lower returns) associated with residential mortgages. However the low profitability also reflected lingering interest rate exposures and a failure to cut costs in a much more competitive mortgage market.
Additionally, there has been a structural change in the demand for mortgages from S&Ls. In the two decades after WW II, the assets of the thrifts grew at a rate of 15% per year, three times as fast as those of the banks. But many in the post-war generations used banks or other lenders — insurers such as Prudential, or mortgage brokers such as Countrywide Credit to finance their home purchases.
In response to these threats, deregulation of the thrift industry has allowed thrifts to offer a diversified product base, including auto loans, checking accounts, and mutual funds. Ironically, this deregulation has given thrifts, in some products, more freedom than banks. For example, S&Ls do not face the same geographic limitations as banks when it comes to selling insurance. National banks may sell credit-related, fixed annuities without geographic restrictions. However, liability, casualty, automobile, life, health and accident insurance must also be sold in towns of 5,000 or less where the bank has a branch. Thrifts face no such restrictions. Thrifts also have the most liberal branching rights of all federal depository institutions, and virtually unlimited holding company activities. This means that any business can own a thrift through a holding company. This is not the case with banks; non-banks cannot own banks.
The answer for many to this situation is to merge the bank and thrift charters and their respective insurance funds. But before that happens, there has to be a broad rewrite of banking legislation. Both banks and thrifts now appear to favor such reforms. After fighting for decades, the two industries are both lobbying for legislation that could merge the bank and thrift charters. Their efforts have become mired, however, in the ongoing discussion on the repeal of the 1934 Glass-Steagall Act separating commercial and investment banking.
6 E.J. Kane, in his book The S&L Insurance Mess (Washington, The Urban Institute Press, 1990), estimated that if in 1983 all insolvent thrifts had been closed down, the cost to the FSLIC would have been only about $25 billion.
Mortgage-Backed Securities
Aside from credit risk, the main characteristic that separates mortgages from other types of fixed-income securities is the degree of uncertainty associated with the cash flows of principal repayment and interest. Even Treasury or corporate callable bonds have provisions that prohibit calling or refunding the bond prior to a certain date. Also, the bondholder can expect that the issuer will not exercise the call option when the coupon rate is less than the current market interest rate. This is not the case with fixed-rate mortgages.
The uncertainty about the cash flows in a mortgage comes from the call option that the borrower holds to prepay (call) part or all of the mortgage at any time. Any payment that is made in excess of the regularly scheduled principal prepayment is considered a prepayment. While the prevailing interest rates are an important determinant affecting the homeowner’s decision to prepay, other factors come into play (such as sale of the home, death of the homeowner, personal tax considerations, or misinformation), resulting in unexpected prepayments even when the prevailing mortgage rate is greater than the loan rate.
An investor in a pass-through security is exposed to the total prepayment risk associated with the mortgage pool. There are two types of prepayment risk: contraction risk and extension risk. Contraction risk is the risk of prepayments when interest rates have declined, forcing the investor to reinvest the cash flow at lower (market) rates of return. Extension risk is the slowdown in prepayments when interest rates rise, resulting in reduced cash flow that can be reinvested at the higher interest rates.
In 1983, Freddie Mac introduced a security called a collateralized mortgage obligation (or CMO). CMOS can be backed by pools of agency pass-throughs or by whole loans. In a CMO, the distribution of principal repayment (both scheduled and prepayments) is done on a prioritized basis to different tranches, so as to redistribute prepayment risk among the different tranches. In that way, each tranche’s prepayment schedule can be predicted much more accurately, reducing prepayment risk.
In a plain vanilla CMO, only one tranche would receive principal repayment until it was completely paid off. Then the second tranche would start receiving principal repayments until this tranche was completely paid off. The CMO tranches are usually labeled alphabetically, with the last tranche to pay off always known as the Z tranche. Other kinds of CMOs, such as the planned amortization class (PAC) bonds, or the Targeted Amortization class (TAC) bonds, are designed to have more precise estimates of prepayments, or to protect investors only against extension or contraction risk, but not both.
In 1986, another type of derivative MBS was introduced by Fannie Mae. This security is known as the stripped MBS. In a stripped MBS, all of the interest from the pass-through is distributed to one tranche, known as the Interest Only class, or IO “strip”, while all the principal was assigned to the Principal Only class, or PO “strip”. Stripped MBS allowed investors to create synthetic securities with risk-return profiles not available in standard CMOs, and provided institutional investors with instruments that could be used to more effectively hedge their liabilities.
1. What are the similarities and differences of the problems encountered by the S&L industry and the more recent problems of the mortgage / housing industry?
2. The US government has created several agencies – the FHA, FHLB, Fannie Mae, Freddie Mac, Ginnie Mae – to provide assistance to the housing market and essentially took over Fannie and Freddie in mid-2008 to assure that it could continue to operate. Why has this sector attracted so much attention and assistance?
3. What policy actions and changes to the marketplace undermined a solid position of savings and loans in the 1970s? Could the effects of these changes have been anticipated and foreseen?
4. Many actions were taken by Congress and regulators to try to address the weakening position of savings and loans. Why were these not effective?