Mortgage-backed security

The securitization of mortgages in the 1970s had the advantage of providing more capital for housing at a time when the demographic bulge of baby boomers created a housing shortage and inflation was undermining a traditional source of housing funding, the savings and loan associations (or thrifts), which were limited to providing uncompetitive 5.75% interest rates on savings accounts and consequently losing savers' money to money market funds. Unlike the traditional localized, inefficient mortgage market where there might be a shortage or surplus of funds at any one time, MBSs were national in scope and regionally diversified.[9] Mortgage backed securities helped move interest rate out of the banking sector and facilitated greater specialization among financial institutions.

However, mortgage-backed securities may have "led inexorably to the rise of the subprime industry" and "created hidden, systemic risks". They also "undid the connection between borrowers and lenders". Historically, "less than 2% of people lost their homes to foreclosure", but with securitization, "once a lender sold a mortgage, it no longer had a stake in whether the borrower could make his or her payments."[10]

As part of the New Deal following the Great Depression, the US federal government created the Federal Housing Administration (FHA) with the National Housing Act of 1934 to assist in the construction, acquisition, and rehabilitation of residential properties.[14] The FHA helped develop and standardize the fixed-rate mortgage as an alternative to the balloon payment mortgage by insuring them, and helped the mortgage design garner usage.[15]

In 1938, the government also created the government-sponsored corporation Fannie Mae to create a liquid secondary market in these mortgages and thereby free up the loan originators to originate more loans, primarily by buying FHA-insured mortgages.[16] As part of the Housing and Urban Development Act of 1968, Fannie Mae was split into the current Fannie Mae and Ginnie Mae to support the FHA-insured mortgages, as well as Veterans Administration (VA) and Farmers Home Administration (FmHA) insured mortgages, with the full faith and credit of the US government.[17] In 1970, the federal government authorized Fannie Mae to purchase private mortgages—that is, those not insured by the FHA, VA, or FmHA, and created Freddie Mac to perform a role similar to that of Fannie Mae.[17] Ginnie Mae does not invest in private mortgages.

Ginnie Mae guaranteed the first mortgage pass-through security of an approved lender in 1968.[18] In 1971, Freddie Mac issued its first mortgage pass-through, called a participation certificate, composed primarily of private mortgages.[18] In 1981, Fannie Mae issued its first mortgage pass-through, called a mortgage-backed security.[19] In 1983, Freddie Mac issued the first collateralized mortgage obligation.[20]

In 1960 the government enacted the Real Estate Investment Trust Act to allow the creation of the real estate investment trust (REIT) to encourage real estate investment, and in 1977 Bank of America issued the first private label pass-through.[21] In 1983 the Federal Reserve Board amended Regulation T to allow broker-dealers to use pass-throughs as margin collateral, equivalent to over-the-counter non-convertible bonds.[22] In 1984 the government passed the Secondary Mortgage Market Enhancement Act to improve the marketability of private label pass-throughs,[21] which declared nationally recognized statistical rating organization AA-rated mortgage-backed securities to be legal investments equivalent to Treasury securities and other federal government bonds for federally charted banks (such as federal savings banks and federal savings associations), state-chartered financial institutions (such as depository banks and insurance companies) unless overridden by state law before October 1991 (which 21 states did[23]), and Department of Labor–regulated pension funds.[24]

The Tax Reform Act of 1986 allowed the creation of the tax-exempt real estate mortgage investment conduit (REMIC) special purpose vehicle for the express purpose of issuing pass-throughs.[25] The Tax Reform Act significantly contributed to the savings and loan crisis of the 1980s and 1990s that resulted in the Financial Institutions Reform, Recovery and Enforcement Act of 1989, which dramatically changed the savings and loan industry and its federal regulation, encouraging loan origination.[26][27]

Low-quality mortgage-backed securities backed by subprime mortgages in the United States caused a crisis that played a major role in the 2007–08 global financial crisis. By 2012 the market for high-quality mortgage-backed securities had recovered and was a profit center for US banks.[28]

The secondary mortgage market is the market where a network of lenders sell, and investors buy, existing mortgages or MBS. A large percentage of newly originated mortgages are sold by their originators into this large and liquid market where they are packaged into MBS and sold to public and private investors, including Fannie Mae, Freddie Mac, pension funds, insurance companies, mutual funds and hedge funds.

Because of the long-term nature of mortgages, the secondary market is an essential factor in maintaining lender liquidity. The infusion of capital from investors provides mortgage lenders such as banks, thrifts, mortgage bankers and other loan originators with a market for their loans. In addition to providing liquidity and increasing overall efficiency, the secondary market can smooth out geographic credit disparities. However, in some respects, particularly where subprime and other riskier mortgages are involved, the secondary mortgage market may exacerbate certain risks and volatility.[4]

TBAs—short for "to-be-announced" securities—involve a special type of trading of mortgage-backed securities. TBAs are the most liquid and important secondary mortgage market, with volume in the trillions of dollars annually.[38] TBAs are traded by MBS traders with notional amounts. There are settlement days when the traders have to make good on their trades. At that time, they choose fractions from various pools to make up their TBA. Only agency mortgage-backed securities trade in the TBA market.[38] "In a TBA transaction, the parties agree on a price for delivering a given volume of Agency Pass-Through Mortgage-Backed Securities at a specified future date. The distinguishing feature of a TBA transaction is that the actual identity of the securities to be delivered at settlement is not specified on the date of execution (“Trade Date”). Instead, the parties to the trade agree on only five general parameters of the securities to be delivered: issuer, mortgage type, term, coupon, and month of settlement."[39]

TBAs are critical in determining the ultimate interest rates that mortgage borrowers pay, since mortgage originators can "lock in" rates and use TBAs to hedge their exposure. TBAs are also used to hedge many non-TBA eligible mortgage products, such as hybrid ARMs and non-agency mortgages.[38]

In Europe there exists a type of asset-backed bond called a covered bond, commonly known by the German term Pfandbriefe. Covered bonds were first created in 19th-century Germany when Frankfurter Hypo began issuing mortgage covered bonds. The market has been regulated since the creation of a law governing the securities in Germany in 1900. The key difference between covered bonds and mortgage-backed or asset-backed securities is that banks that make loans and package them into covered bonds keep those loans on their books. This means that when a company with mortgage assets on its books issues the covered bond, its balance sheet grows, which would not occur if it issued an MBS, although it may still guarantee the securities payments.

The weighted-average maturity (WAM) and weighted average coupon (WAC) are used for valuation of a pass-through MBS, and they form the basis for computing cash flows from that mortgage pass-through. Just as this article describes a bond as a 30-year bond with 6% coupon rate, this article describes a pass-through MBS as a $3 billion pass-through with 6% pass-through rate, a 6.5% WAC, and 340-month WAM. The pass-through rate is different from the WAC; it is the rate that the investor would receive if he/she held this pass-through MBS, and the pass-through rate is almost always less than the WAC. The difference goes to servicing costs (i.e., costs incurred in collecting the loan payments and transferring the payments to the investors.)

To illustrate these concepts, consider a mortgage pool with just three mortgage loans that have the following outstanding mortgage balances, mortgage rates, and months remaining to maturity:

The weighted-average maturity (WAM) of a pass-through MBS is the average of the maturities of the mortgages in the pool, weighted by their balances at the issue of the MBS. Note that this is an average across mortgages, as distinct from concepts such as weighted-average life and duration, which are averages across payments of a single loan.

The weightings are computed by dividing each outstanding loan amount by total amount outstanding in the mortgage pool (i.e., $900,000). These amounts are the outstanding amounts at the issuance or initiation of the MBS. The WAM for the above example is computed as follows:

WAM = (22.22% × 300) + (44.44% × 260) + (33.33% × 280) = 66.66 + 115.55 + 93.33 = 275.55 months

Another measure often used is the Weighted-average loan age.

The weighted-average coupon (WAC) of a pass-through MBS is the average of the coupons of the mortgages in the pool, weighted by their original balances at the issuance of the MBS. For the above example this is:

WAC = (22.22% × 6.00%) + (44.44% × 6.25%) + (33.33% × 6.50%) = 1.33% + 2.77% + 2.166% = 6.277%

Pricing a "vanilla" corporate bond is based on two sources of uncertainty: default risk (credit risk) and interest rate (IR) exposure.[42] The MBS adds a third risk: early redemption (prepayment). The number of homeowners in residential MBS securitizations who prepay increases when interest rates decrease. One reason for this phenomenon is that homeowners can refinance at a lower fixed interest rate. Commercial MBS often mitigate this risk using call protection.[43]

Since these two sources of risk (IR and prepayment) are linked, solving mathematical models of MBS value is a difficult problem in finance. The level of difficulty rises with the complexity of the IR model and the sophistication of the prepayment IR dependence, to the point that no closed-form solution (i.e., one that could be written down) is widely known. In models of this type, numerical methods provide approximate theoretical prices. These are also required in most models that specify the credit risk as a stochastic function with an IR correlation. Practitioners typically use specialised Monte Carlo methods or modified Binomial Tree numerical solutions.

Theoretical pricing models must take into account the link between interest rates and loan prepayment speed. Mortgage prepayments are usually made because a home is sold or because the homeowner is refinancing to a new mortgage, presumably with a lower rate or shorter term. Prepayment is classified as a risk for the MBS investor despite the fact that they receive the money, because it tends to occur when floating rates drop and the fixed income of the bond would be more valuable (negative convexity). In other words, the proceeds received would need to be reinvested at a lower interest rate.[8] Hence the term prepayment risk.

Professional investors generally use arbitrage-pricing models to value MBS. These models deploy interest rate scenarios consistent with the current yield curve as drivers of the econometric prepayment models that models homeowner behavior as a function of projected mortgage rates. Given the market price, the model produces an option-adjusted spread, a valuation metric that takes into account the risks inherent in these complex securities. [44]

There are other drivers of the prepayment function (or prepayment risk), independent of the interest rate, such as:

• economic growth, which is correlated with increased turnover in the housing market
• home prices inflation
• unemployment
• regulatory risk (if borrowing requirements or tax laws in a country change this can change the market profoundly)
• demographic trends, and a shifting risk aversion profile, which can make fixed rate mortgages relatively more or less attractive

The credit risk of mortgage-backed securities depends on the likelihood of the borrower paying the promised cash flows (principal and interest) on time. The credit rating of MBS is fairly high because:

1. Most mortgage originations include research on the mortgage borrower's ability to repay, and will try to lend only to the creditworthy. An important exception to this is "no-doc" or "low-doc" loans.
2. Some MBS issuers, such as Fannie Mae, Freddie Mac, and Ginnie Mae, guarantee against homeowner default risk. In the case of Ginnie Mae, this guarantee is backed with the full faith and credit of the US federal government.[45] This is not the case with Fannie Mae and Freddie Mac, but these two entities have lines of credit with the US federal government; however, these lines of credit are extremely small compared to the average amount of money circulated through these entities in one day's business. Additionally, Fannie Mae and Freddie Mac generally require private mortgage insurance on loans in which the borrower provides a down payment that is less than 20% of the property value.
3. Pooling many mortgages with uncorrelated default probabilities creates a bond with a much lower probability of total default, in which no homeowners are able to make their payments (see Copula). Although the risk neutral credit spread is theoretically identical between a mortgage ensemble and the average mortgage within it, the chance of catastrophic loss is reduced.
4. If the property owner should default, the property remains as collateral. Although real estate prices can move below the value of the original loan, this increases the solidity of the payment guarantees and deters borrower default.

If the MBS was not underwritten by the original real estate and the issuer's guarantee, the rating of the bonds would be much lower. Part of the reason is the expected adverse selection against borrowers with improving credit (from MBSs pooled by initial credit quality) who would have an incentive to refinance (ultimately joining an MBS pool with a higher credit rating).

Because of the diversity in MBS types, there is wide variety of pricing sources. In general, the more uniform or liquid the MBS, the greater the transparency or availability of prices.[38] Most traders and money managers use Bloomberg and Intex to analyze MBS pools and more esoteric products such as CDOs, although tools such as Citi's The Yield Book, Barclays POINT, and BlackRock's AnSer are also prevalent across Wall Street, especially for multi–asset class managers. Some institutions have also developed their own proprietary software. Tradeweb is used by the largest bond dealers (the "primaries") to transact round lots ($1 million and larger).

Complex structured products tend to trade less frequently and involve more negotiation. Prices for these more complicated MBSs, as well as for CMOs and CDOs, tend to be more subjective, often available only from dealers.[38]

For "vanilla" or "generic" 30-year pools (issued by Fannie Mae, Freddie Mac, or Ginnie Mae) with coupons of 3.5%–7%, one can see the prices posted on a TradeWeb screen by the primaries called To Be Allocated (TBA). This is due to the actual pools not being shown. These are forward prices for the next 3 delivery months since pools have not been cut; only the issuing agency, coupon, and dollar amount are revealed. A specific pool whose characteristics are known would usually trade "TBA plus {x} ticks" or a "pay-up", depending on characteristics. These are called "specified pools", since the buyer specifies the pool characteristic he/she is willing to "pay up" for.

The price of an MBS pool is influenced by prepayment speed, usually measured in units of CPR or PSA. When a mortgage refinances or the borrower prepays during the month, the prepayment measurement increases.

If an investor has acquired a pool at a premium (>100), as is common for higher coupons, then they are at risk for prepayment. If the purchase price was 105, the investor loses 5 cents for every dollar prepaid, which may significantly decrease the yield. This is likely to happen as holders of higher-coupon mortgages can have a larger incentive to refinance.

Conversely, it may be advantageous to the bondholder for the borrower to prepay if the low-coupon MBS pool was bought at a discount (<100). This is due to the fact that when the borrower pays back the mortgage, he does so at "par". If an investor purchases a bond at 95 cents on the dollar, as the borrower prepays the investor gets the full dollar back, increasing their yield. However, this is less likely to occur, as borrowers with low-coupon mortgages have lower, or no, incentives to refinance.

The price of an MBS pool is also influenced by the loan balance. Common specifications for MBS pools are loan amount ranges that each mortgage in the pool must pass. Typically, high-premium (high-coupon) MBSs backed by mortgages with an original loan balance no larger than $85,000 command the largest pay-ups. Even though the borrower is paying an above market yield, he or she is dissuaded from refinancing a small loan balance due to the high fixed cost involved.

Low Loan Balance: < $85,000
Mid Loan Balance: $85,000–$110,000
High Loan Balance: $110,000–$150,000
Super High Loan Balance: $150,000–$175,000
TBA: > $175,000

The plurality of factors makes it difficult to calculate the value of an MBS security. Often market participants do not concur, resulting in large differences in quoted prices for the same instrument. Practitioners constantly try to improve prepayment models and hope to measure values for input variables implied by the market. Varying liquidity premiums for related instruments and changing liquidity over time make this a difficult task. One factor used to express price of an MBS security is the pool factor.