Two distinct but related legal issues have emerged in the housing markets over the past few weeks that, if left unresolved, could undermine future private-sector participation in the U.S. mortgage markets. First, faulty documentation practices threaten to void thousands of foreclosures, or have them delayed by the courts. The results would be billions of dollars of losses for banks seeking to seize and sell the underlying collateral on mortgage loans that had gone bad. Second, investors in mortgage-backed securities (MBS), including the Federal Reserve Bank of New York, recently wrote to Bank of America (BAC), demanding that it repurchase $47 billion in mortgage loans securitized and sold by Countrywide (which was acquired by BoA in 2008). The contention is that the loans sold into MBS pools were “marketed with misstatements about quality” that is tantamount to fraud. Specifically, the investors charge that Countrywide relied on false appraisals of the properties financed by those mortgages.
This legal uncertainty layers new risks on top of those that had already made mortgage investing particularly toxic. As explained in a seminal 2000 paper (pdf) in the academic journal Econometrica, mortgages are particularly risky instruments for investors for four reasons:
(1) Mortgages share with all fixed-income instruments the risk that interest rates will increase and cause the market value of the security to fall;
(2) Mortgages have an embedded “call” option that allows the mortgage borrower to refinance the mortgage whenever interest rates fall below the coupon rate on the mortgage. When rates fall to 5%, for example, borrowers with a 6% mortgage are generally free to refinance their existing loan and save on the lower interest rate. As a result, investors in the mortgage are deprived of the normal benefits of falling interest rates since the increase in the value of the mortgage above par is generally truncated once homeowners choose to refinance into new mortgages.
(3) Mortgages have an embedded “put” option that allows the mortgage borrower to stop making payments on the mortgage as soon as the price of the house falls below the unpaid principal balance on the mortgage. In most states, mortgages are non-recourse in law or in practice. This means borrowers with a $300,000 loan on a house worth $200,000 can voluntarily vacate the property and save $100,000 with no concern that the bank can seize assets to make up for the deficiency.
(4) Finally, the households electing to exercise these options are heterogeneous and make decisions that are difficult to forecast. Few households exercise their “call option” to refinance their mortgage optimally and some never refinance even if doing so could generate significant savings. Other households exercise their option to prepay their mortgages even when the option is not “in the money.” For example, a household with a 5.5% mortgage might move from New York to Denver at a time when mortgage rates are 6.25% and the market value of the mortgage is below par. Even more difficult to forecast is the decision to exercise the “call option” and walk away from a mortgage. While few borrowers stop making payments as soon as their home is worth less than the mortgage, once the home falls 15% below the value of the mortgage with no signs that housing prices will rebound, defaults increase exponentially.
For many years, analysts focused on the interest rate and prepayment risks (risks 1 and 2 above). The concern about Fannie Mae (FNMA.OB) and Freddie Mac (FMCC.OB) was that both GSEs could be rendered insolvent by large increases (decline in market value of mortgage assets) or decreases (massive wave in prepayments leaving assets worth less than liabilities) in interest rates. It was not until 2006 that analysts focused more on credit risk (risk 3 above). An excellent 2006 paper (pdf) from the Federal Reserve Bank of Philadelphia explained how the existence of the put option could lead to a massive spike in delinquencies if house prices fell by 10% or 20%. As is now clear, no mortgage investor was prepared for the volume of options that would be exercised if prices plunged more than 20%. Had price declines of this magnitude been seriously contemplated by lenders – or the investors who ultimately funded mortgages – it’s unlikely the mortgage credit boom would have ever materialized.
More significantly, few investors were prepared for the heterogeneous exercise of the options written to borrowers. In counties and states with large numbers of investment properties, like Las Vegas, borrowers tended to walk away as soon at the option was “in the money” since they didn’t live in the house and it made no sense to continue making payments on a loss-making mortgage. As a result, foreclosure sales account for nearly 3 of every 5 home sales in Nevada. In other states, the share of investment properties was lower and required price declines more substantial to lead borrowers to walk away from properties. In other cases, the difference in option exercise was livability of the community. For example, borrowers who acquired homes in far-out suburbs with long commute times may have been more inclined to walk away from homes that plummeted in value than borrowers living in homes located in close-in suburbs that experienced similar price declines.
Financial theory would predict that the value of the call and put options embedded in mortgages would be reflected in the interest rates. But the theory doesn’t match with reality, largely because of government involvement in the market through the GSEs and irrationally optimistic views about the improbability that house prices would or could ever fall substantially on a national basis. The repricing of these options has made private sector lenders uncompetitive in the market. The federal government-backed lenders accounted for 97% of mortgage loans in the first part of 2010 largely because they were the only entities willing to fund mortgages at 4.5%.
Click to enlarge
In addition to the repricing of embedded mortgage options, investors now have to consider that the fundamental mechanism through which funds are channeled to mortgage borrowers could be broken or, at the very least, dysfunctional. The risks associated with the foreclosure process are fundamental to mortgage investing because the house is what ultimately secures the loan. The mortgage crisis was initially focused on loans to borrowers with bad credit histories. But if the foreclosure process works efficiently, the losses on these loans should be trivial as the investors take possession of the homes, sell them, and receive the cash they were owed. It is only when the house value declines precipitously or when the investor cannot take possession of the home that the losses for investors grow far beyond expectation. Concern about foreclosure processes could add an additional point or more to the yield private investors would demand to fund mortgages.
This uncertainty also relates directly to the second issue: concern about the appraisal practices and loan documentation. There are obvious information asymmetries between mortgage originators and investors concerning the credit quality of the borrower, the value of the house, and other factors critical to the performance of the mortgage. If investors think the loans sold into pools are systematically deficient in terms of due diligence, the price of mortgage credit will be even higher. If investors think that a house appraised for $250,000 according to the documentation that accompanies MBS issuance could be worth as little as $200,000, they will demand additional compensation to account for the potential losses on a $200,000 mortgage if the borrower defaults on the loan. Having to account for the unreliability of the data accompanying MBS will dramatically increase mortgage yields and, in the limit, close the securitization market entirely.
All of these problems can be overcome, of course, through a government guarantee. When a government agency puts its stamp on an MBS, concern about credit risk, foreclosure, and the integrity of data reporting all disappear as any problems in these areas are assumed by taxpayers. Unfortunately, the cost of this approach is quite high, as everyone is now well aware, given the combined $226 billion in losses (pdf) recognized at Fannie and Freddie in just two years.
Are policymakers prepared to turn the mortgage markets over to private investors if it means mortgage rates approximate the 7.2% demanded for high-yield corporate loans (see Merrill Lynch index above), in contrast to the current 4.19% rates offered by the government agencies? If policymakers want a private mortgage finance system at rates comparable to those available today, they may very well have to embrace a number of provisions that are often labeled anti-consumer, such as eliminating the free prepayment option, making mortgages fully recourse loans, expediting the foreclosure process, or requiring even more personal attestation on the part of borrowers.
When the government artificially suppresses prices, simply getting out of the way and allowing “the market to work” does not always lead to perfect outcomes.