Mortgage Backed Securities
When an individual or family purchases real estate the typically borrow from a bank or other lender. The bank (or lender) will issue a mortgage to finance the purchase of the home. In return, the borrower will make fixed monthly payments comprised of interest (based on the balance of the mortgage) and principal (that equals the difference between the fixed payment and the interest component). Monthly payments are made over a set time period, typically 30 years, until the mortgage is paid off and the principal balance equals zero. Typically, since most property is not held for the length of the mortgage, mortgages are paid off early (after 10 years for example when the property is sold). Other mortgages may be prepaid when interest rates drop and a new mortgage is originated at a lower rate of interest.
After a mortgage is originated by a bank or lender, many of these mortgages are sold off by the lender in a secondary market. Buyers are these mortgages are either investment bankers or a federal agency known as the Federal Home Loan Mortgage Corporation (FHLMC or Freddie Mac). By pooling together multiple mortgages, the investment banker or FHLMC will issue what is generally known as mortgage backed securities. Mortgage backed securities are sold as bonds and their coupon payments and par value are derived from the original mortgages from which they are created.
For example, assume that you take out a $200,000 mortgage from the local bank to buy your dream home. Your bank then sells your mortgage in the pass-through market to an investment banking firm that groups together mortgages in order to sell mortgage backed securities to buyers such as a pension fund. In return your bank receives the value of the mortgage, $200,000, that it can again lend out to a different borrower Although you still make your monthly payment to the bank that originated your mortgage, your interest and principal payment is passed-through to the buyer of the mortgage backed security. You bank will receive a small processing fee, the investment banker takes a cut by offering a coupon on the bonds that is slightly less than the interest payments on the mortgage, and the buyer gains a slightly higher return that a corresponding 10-year Treasury note. Since most home mortgages are guaranteed by the Federal Housing Administration (FHA), the risk of default is essentially zero.
Mortgage backed securities are a class of derivatives since the owner of these bonds derives the return from the payments on mortgages, but does not actually own the mortgage itself.
Collateralized Mortgage Obligations
The average mortgage has a 7 year life span. However, some investors want CMOs with a shorter time to maturity and others prefer a security with a longer time to maturity. CMOs - collateralized mortgage obligation, a security whose value is created out of bundles of residential mortgages as well as securities derived from them based on the rate at which homeowners repay mortgages. CMOs differ from other mortgage-backed securities in that they are issued in several different maturity classes (called tranches) based on a projected schedule for repaying the mortgage loans.
Over time, more exotic use of basic Fannie Mae's developed into a market known as collateralized mortgage obligations or CMOs. Since mortgage backed securities are bonds backed by pools of residential mortgages they may be highly unpredictable due to the varying repayment patterns of homeowners who often prepay mortgages. Buyers of mortgage backed securities found this attribute of mortgage backed securities to be less than desirable.
For example, a manager of a pension fund may buy a group of mortgage backed securities with a 10 year maturity. The money manager desire to balance the income from the securities with payment obligations to pensioners. But if market interest rates fall and mortgage refinancing increases the mortgage bonds may begin to mature early. As a result, the pension fund manager must reinvest the money in securities that reflect the lower market interest rates and thus have a lower yield. During the 1980s, Wall Street attempted to reduce the prepayment uncertainty by breaking down these securities into different components.
The FHLMC or investment bankers group together mortgages into similar attributes such as the interest rate on the mortgage, amount of the mortgage, prepayment risk of the mortgage (e.g. younger home buyers in California may be more likely to move and sell their home after a few years than a middle-aged buyer in Michigan), and other possible attributes of the buyer and mortgage.
The key feature of the CMO structure is the prioritization of the cash flows from a pool of mortgages among several different classes of bondholders. This mechanism creates a series of bonds with varying maturities that appeal to different investors. The first CMOs were issued in 1983 by the FHLMC.
The cash flows received from a CMO are primarily determined by the payments from the underlying mortgage principal and interest payments. Unlike more traditional fixed-income securities whose cash flows can be predicted with relative certainty, the cash flows from a mortgage are subject to variation. The uncertainty of these cash flows results from borrowers having the option to prepay their mortgages at any time prior to final maturity. These prepayments can occur for many reasons, as CMOs are similar to other mortgage backed securities. For example, when homeowners move or refinance their loans to take advantage of lower interest rates, prepayments will occur. Prepayments, therefore, tend to increase as mortgage loans age and as interest rates fall.
The CMOs issued by the Federal Home Loan Mortgage Corporation (Freddie Mac), for example, are generally structured with 3-6 classes (tranches) of bonds with different maximum maturities. Non-FHLMC issues also have more than three tranches. All mortgage payments, including scheduled amortization of mortgage principal, prepayments, and interest payments, are passed through to investors according to the provisions of the original specifications of that tranch.
CMOs enable investors to manage aid, in most cases reduce their exposure to the prepayment risk. Because of this, CMOs have greater value and apparently lower yield requirements for the investor than plain-vanilla mortgage backed securities. CMOs with different times to maturity display a positively sloped yield curve. CMOs offer higher yields than other investments of similar quality and maturity as compensation for somewhat variable and uncertain average lives.
Holders of the longer maturity classes of CMOs face a lower risk of the bond maturing early when mortgages are prepaid than with a simple mortgage backed security. Although we won't go into the details, the reason deals with the allocation of principal payments among the different tranches of bonds in a group of CMOs of different maturities.
Initially, CMOs were given 2 slices, one offering a shorter maturity and the other a longer. Now a modern CMO may have as many as 20 or more slices of varying times to maturity and other attributes such as risk of default.
For example, recent CMO issuers have added another new feature to the CMO security: a compound interest or accrual bond (also referred to as a Z bond). Investors in this class, typically the class with the longest maturity, receive neither principal or interest payments until all of the preceding classes of the bond have been retired. The coupon income foregone is added to the outstanding principal balance of the bonds. This feature provides, in effect, automatic reinvestment of the principal balance at the coupon rate. When all prior tranches are fully retired, the compound interest bond begins to pay principal and interest.
Z bonds offer long-term investors high yield and quality, a long (but uncertain) duration and automatic reinvestment of cash flows for a number of years. They provide yields that are often higher than those available from typical fixed-income investments, lower-quality bonds, private placements and traditional mortgage products. Z bonds, as most CMO issues, are generally AAA-rated and/or backed by agency-guaranteed mortgage securities.
Toxic Waste
During the 1960s and 1970s there was a group of bond dealers known as the Memphis Bond Daddies. The Bond Daddies were Tennessee outfits notorious for their unscrupulous methods of selling tax-exempt bonds to the elderly. Eventually closed down by regulators, many of the Bond Daddies would reappear in Houston, Texas in the early 1990s. There role was to dispose of the toxic waste generated as CMOs were sliced into more and increasingly complex tranches. As the number of tranches offered by investment banking firms grew so did the complexity of the CMOs offered as each become more specific in its characteristics.
To improve the ability of investors to buy CMOs with specific time to maturity and risk, CMOs were further enhanced to separate the interest and principal payments into interest only CMOs (IO) and principal only (PO) securities.
Unlike conventional bonds, the price of an IO has a positive relationship to interest rates. for example, as interest rates fall, IO prices fall due to increasing prepayments, reducing the flow of interest payments. IOs offer a hedge against conventional bonds where prices are inversely related to interest rates. As with conventional bonds the prices of PO's move inversely to interest rates. As interest rates fall, prepayments accelerate implying that investors get their principal back earlier.
Another variation created is known as floating-rate CMOs, where coupons are reset monthly to an interest rate index (usually the fed funds rate). A further tranche from Floating-rate CMOs was created as an inverse floater. Inverse floaters and closely related inverse IOs are CMOs whose coupons float inversely to a interest rate index.
If interest rates are falling the effect on inverse floaters and inverse IOs works as follows:
- Since they are mortgage bonds, their prices rise as market interest rates fall.
- Unlike typical mortgage backed securities and bonds, their coupon is not fixed, but moves inversely to market interest rates. As interest rates fall, the coupon rises.
In the early 1990s, interest rates were falling due to weak economic conditions. Money managers of pension funds, municipal governments, colleges and other institutions are often limited in the type of investments they can undertake with the funds of their institution. With the falling market interest rates present in 1992 and 1993, the yield from buying traditional low-risk bonds was relatively low. Since most of these money managers were not allowed to buy higher-risk and yield junk bonds, some searched for the highest possible yield in bonds with a low-default rating.
Money managers such as Robert Citeron of Orange County, the money manager of Odessa College (Odessa, Texas), the Minneapolis Symphony Orchestra, the Piper Jaffery manager of the fixed-income mutual fund and many others directed much of the portfolio they managed into CMOs, and especially inverse floaters and inverse IOs.
Kidder Peabody and other investment banking firms that were underwriting new CMO offerings created inverse floater and inverse IOs in the early 1990s to offer a higher return during periods of falling interest rates. As noted, when interest rates are falling both the price and coupon of these two CMO tranches are rising, given an extra kick to the rate of return earned on these bonds. And since these bonds were CMOs, they were considered to have minimal default risk.
The problem that the CMO wizards at Kidder Peabody and other investment bankers realized was the unknown behavior of these two tranches when interest rates increased. When CMOs with 20 tranches were created out of a pool of mortgages, 90% (18 out of the 20 tranches) of the hybrid CMOs are fairly predictable in there response to changes in market conditions, relatively safe and desirable to sophisticated investors looking for a little higher return with little additional risk.
The risk was concentrated into the remaining two tranches - the inverse floaters and inverse IOs.
The investment bankers issuing the more complicated CMOs created a dependent system - in order to create the safer, easily marketable 90%, the risk needed to be concentrated in the remaining 10% (inverse floaters and inverse IOs) which were known as the "toxic waste".
So the large investment banking firms had a huge market for the 90% docile CMOs but needed to dispose of the toxic waste in order to create them. For example, if the investment banking firm acquired $500 million worth of mortgages, they would generate $450 million in safe CMOs. The risk would be concentrated in the remaining $50 million worth of CMOs - the inverse floaters and inverse IOs - the byproduct of creating the other 18 tranches.
The investment banking firms needed to dispose of this toxic waste since otherwise, they would either lose their value of $50 million or be stuck with $50 million in securities that they realized could act lethally when interest rates increased.
The large investment banking firms did not want to risk the legal liability of selling toxic waste, so the disposal and sales were farmed out to about a dozen smaller securities firms, mostly based in Houston, who used high pressure tactics to sell to unwary institutional investors. For example, Kidder Peabody sold over 60% of its toxic waste through Houston based firms such as Government Securities Corp. (GSC), and Westcap.
During the early 1990s, falling interest rates made the extra return on inverse floaters and inverse IOs look very attractive to some money managers who were looking for a way to look better and smarter than their peers who had a lower rate of return on the assets in their portfolio. Using high pressure tactics and repeated calling of basically ignorant clients the toxic waste was easily disposed by the Houston firms. With large margins on sales, successful salesmen earned over $1m a year.
But then it was 1994 and the Fed decided that the economy was running too hot and the brakes needed to be applied. Seven times the Fed raised interest rates that year, decimating the market for inverse floaters and IOs. With rising rates, not only are the prices of these two tranches inversely related to interest rates, but so is their coupon payments. Falling coupon payments further plummeted the market price of these securities.
The money managers who briefly rode inverse floaters and IO;'s to higher returns, were crushed in 1994. Being highly leveraged in these securities, they needed to sell off part of their portfolio at large loss to cover expenses. For example, Robert Citeron, the money manager for Orange County, California, had spent a good part of 1992 and 1993 boasting of his genius when it came to investing since he was earning a higher return than his counterparts in other municipalities across the nation. But as security prices crashed in 1994, he had to sell the portfolio at a huge loss in order to fund daily operating expenses for Orange County. By the end of the year Orange County was technically bankrupt and had to pay its employees and suppliers with chits and IOUs.
Other money managers lost just as spectacularly. The money manager for Odessa College in Texas lost 50% of its $21.9 million operating fund on inverse IOs in the first half of 1994. As a result, classes and faculty were slashed to minimize the college's budget. City Colleges of Chicago purchased over $250m in risky mortgage derivatives. Money was gambled and lost by the public school district of Vermilion, Ohio, on the shore of Lake Erie. The Shoshone Tribe on the Wind River reservation in WY lost $3.5m in 1994.