The Donkey-Backwards Housing Finance Debate

One of the biggest questions in housing finance is how to revive the securitization market. During the housing boom of the early 2000s, banks earned massive amounts of fee and commission income by packaging mortgages into mortgage-backed securities. These securities were often sliced and diced into CDOs, CDOs squared and what not, in order to further entice investors (and speculators). What happened next is all too well known. The banks, eager to bulk up low-risk fee revenue while offloading lending risk, thought that if writing a lot of mortgage loans was good, then writing a shipload more would be even better. Lending standards dropped, to the point where banks didn't always document borrowers' incomes, as if to avoid learning that they shouldn't extend the loan. There were plenty of times when they shouldn't have, but did anyway.

That insouciance toward prudence dug a very deep grave for investor interest in securitized loans. Today, just about the only mortgage-backed securities that can be sold carry explicit U.S. government guarantees. Housing finance has become a federal program, and today's housing stock enjoys what is effectively a federal price support policy.

Needless to say, taxpayers can't support housing values indefinitely. In the view of bankers and many regulators, securitization must be improved and revived. The two potential improvements most often discussed are (a) requiring the banks that package mortgages to keep some of the lending risk, or (b) improving underwriting standards without requiring underwriting banks to keep some "skin in the game." The first concept is intended to keep the banks honest. But banks holding increased amounts of lending risk also must increase their capital levels. That is likely to lower profits, anathema to their executives suites and also not to the liking of some bank regulators, who seem to equate lower bank profits with greater aggravations for themselves. The second notion--improved underwriting standards--is clearly necessary, but insufficient by itself. Investors aren't prepared to put down their money with just promises of improvement.

The problem is the discussion focuses on what the banks (and regulators) want, not what investors would like. This is donkey-backwards. A revival of a private securitization market depends on the willingness of investors to plunk their cash onto the barrelhead. There used to be a notion in American business that the customer is always right. A little fillip of customer service--i.e., investor protection--needs to be added to the mix.

First, there's the issue of trust. Trust is the true foundation of the financial system. Investors no longer trust the banks at the heart of the securitization process. That's why the only mortgage-backed securities acceptable to investors today bear a federal guarantee. Banks hoping to securitize on their own seem to be viewed as little more than potential scofflaws. Serious regulatory reform--of both banking and the derivatives market--is essential. Consumer protection must be greatly strengthened, lending standards bearing a reasonable resemblance to prudence have to be enforced, and the derivatives market must become much more transparent. But the scope of reform evolving in current legislative proposals may be inadequate to reassure holders of capital.

Second, the securitization market as it existed in the early 2000s ceased to be risk-sensitive. Investors had no effective way to discern that they were buying bags of digestive waste, and banks securitizing loans ceased to care that they were selling the same. The absence of risk sensitivity created grotesque market distortions that resulted in millions of bad loans being made, which may have enriched underwriting banks but also led to the defaults and foreclosures that have been driving down real estate prices.

Risk insensitivity is the problem that the skin-in-the-game requirement is intended to fix. The continued desertification of the securitization market is a signal that not enough is being done. Further product development is required. Perhaps banks should agree to limit investor losses to a predetermined number of cents on the dollar invested. After that, the underwriting bank would bear all losses. The less the investment resembles a pig in a poke, the more likely people will buy. Such a provision would improve the quality of mortgages in the pool, which would benefit investors--and homeowners. Fewer low quality loans would be made. Even if home ownership levels fall, bad loans do not, in the medium (let alone, long) term, increase home ownership. They do, however, drive down real estate values when borrowers default and end up in foreclosure.

Another improvement would be for banks to open up their databases concerning the underlying mortgages to credit rating agencies, and indeed, investors, for analysis. People are more likely to buy if they can kick the tires and lift up the hood. Any competitive issues would be unimportant if all offerings are subject to inspection. Of course, this may make pricing more accurate, or, stated otherwise, fairer to investors. And that's the idea. People will pay a fair price for what they understand, but not a penny for the opaque, black box CDOs of yore. Bank profits would be lower. But the current miserly dialogue about minimizing the extent of improvements to the securitization process may be holding down underwriting banks' costs, at the expense of expunging investor interest.

The bank-centric orientation of reforming the securitization market isn't even leading investors to water, let alone inducing them to drink. However, if banks and regulators would give a nod to the holders of capital who've been taking it on the chin for the last few years, maybe they'd see the phoenix rise from the ashes.