Imagine going into a community bank and getting a mortgage but not telling the lender exactly where the house you want to buy is. You give the lending officer the zip code, and a few other clues about the home, but you walk out of the bank without having given your name or a street address.
Sounds like a scam, right?
Well, that’s how most mortgages are financed in the U.S. Investors in private, residential mortgaged-backed securities (or RMBS) are prevented by law from knowing the address for the mortgages they purchase. Naturally, a bank wouldn’t want to lend a few hundred thousand without knowing this information. We argue in a recent study by Reason Foundation that mortgage investors should be allowed the same privilege. It’s an important part of getting the housing market on a path toward recovery.
The market for mortgage-backed securities collapsed as a result of high foreclosure rates, particularly on subprime mortgages. Investors lost considerable sums on AAA rated mortgage-backed bonds and have been suing rating companies ever since.
Critics of these lawsuits are correct in suggesting that many investors failed to perform due diligence, blindly trusted the ratings system and ignored prospectuses. But even greater investor attention probably wouldn’t have prevented the financial crisis. Since RMBS investors were given only partial details about the properties backing their bonds, and the borrower address wasn’t among them, they could conduct only a limited risk analysis.
Before the mortgage meltdown began in 2007, some had argued that investors shouldn’t have to worry about the exact details of underlying mortgage loans, because the large number of loans in a given portfolio reduced the risk posed by individual borrowers. Clearly, that logic was flawed.
Many investors would prefer to do more research on their own, or work with analytic firms to do so. But ignorance of the borrower’s address and identity is a substantial disadvantage. Perhaps the single-most-important predictor of a mortgage default is the ratio between all mortgage debt on a property and its current value, the so-called combined loan-to-value ratio, or CLTV. Loan-level data provides the CLTV at the time a deal is originated, but this value isn’t updated over the life of the deal.
When prices were soaring at the peak of the bubble, a mortgage in Phoenix or Las Vegas might have had a relatively safe CLTV, even on a mortgage with little down payment. But as prices rapidly declined, many of those homes began to slip “underwater” with a risky CLTV. Without address-level data, it’s impossible to get a precise fix on a mortgage’s value in real time. As a result, it was challenging for investors to understand how rapidly the value of their mortgage-backed bonds was declining.
As we point out in our study, zip codes can contain several thousand properties and may also include large, often heterogeneous areas. For example, single-family homes recently listed on Zillow.com in zip code 20002 (in Northeast Washington) ranged from $250,000 fixers in the Little Trinidad neighborhood to a $2.8 million upscale townhouse on Capitol Hill. Zip codes simply don’t provide enough data to give more than an imprecise estimate of a change in property value.
Lawmakers, regulators and industry leaders want to keep identifying information out of investor reports to protect borrowers’ confidentiality. The language of the Fair Credit Reporting Act of 1970 -- the foundation of consumer credit rights -- is a bit hard to interpret, but giving out borrower addresses to investors probably violates its intent. Doing so also clearly violates a nonessential clause of the proposed Private Mortgage Market Investment Act now pending before the House Financial Services Committee.
Protecting privacy is important. But mortgages are typically recorded in county registers and made available to anyone visiting a county clerk’s office. This is one reason why mortgagors receive refinance offers in the mail: A lender has obtained publicly available mortgage records that he then uses to solicit refinancing business.
If another lender can see that you owe money on your house, why shouldn’t investors in the bonds that back your mortgage see that, too? Clearly some investors value this information because they pay companies such as CoreLogic to match data in RMBS portfolios against these public records. The matching process can be costly and isn’t totally reliable.
It would be cheaper and easier for investors to receive this publicly available information with their mortgage listings. This would encourage more due diligence and give investors more confidence to jump back into the market.
One way to address privacy concerns would be to allow borrowers to choose during the mortgage-origination process whether to allow disclosure of their addresses to investors. In exchange, those who provide authorization could be compensated with reduced closing costs or lower interest rates since their mortgages would be easier to securitize. Privacy would ultimately be the borrower’s choice.
Having access to address-level data wouldn’t solve all the problems in the housing-finance market. But it would make risk analysis substantially more accurate. Combined with a few other reforms -- such as encouraging a common format for RMBS collateral data and creating a Mortgage Underwriting Standards Board to compete with ratings companies -- it would lead to private capital taking back a sizable chunk of credit risk from the government.
And that would be good news for taxpayers. If the mortgage- backed-securities market rebounds, government-backed Fannie Mae (FNMA) and Freddie Mac (FMCC) could be scaled down without simultaneously shutting down housing finance. Taxpayers, homeowners and investors would all win.
(Marc Joffe is a principal consultant at Public Sector Credit Solutions. Anthony Randazzo is director of economic research at the Reason Foundation. The opinions expressed are their own.)
This article first appeared at Bloomberg View on June 3, 2012.