Ethan Penner has an interesting article in today's Wall Street Journal (available for free I think on their Opinion Journal page) on the future of securitization. I discussed the same topic in part 5 of my series "Why Did the Mortgage Crisis Happen?"
The bond markets ended up selling very risky mortgage-backed securities, but there were two sources of the risk. The first was explicit risk: The bondos knew they were dealing in subprime mortgages. The second source was implicit: The loan originators were cutting underwriting corners, lending to even riskier folks than their securitization paperwork implied. The bond markets were partly at fault there, for they were not auditing the underwriting standards, in an environment where the originator has a large incentive to fudge the numbers. Sort of reminds you of American companies importing Chinese toys.
Penner suggests that the securities backed by loan pools only be made of the safest part of the deal, with the originator keeping a good portion of the risk. That's one way to solve the problem, but it ignores one of the good things about securitization: The party with the strongest appetite for risk may not be the party with the most expertise in originating loans. Another viable solution is for the bond buyers to insist on underwriting audits.
Where will we end up? Securitization is vital for the home mortgage market. It's nice for other instruments, such as credit card balances, car loans, and commercial loans, but it's not absolutely necessary in these sectors because banks can borrow for the same term as these loans. That's not possible in the world of 30-year mortgages. But these "unnecessary" securitization sectors are NOT where we've seen the big problems. They are functioning decently.
It all comes back to caveat emptor: Let the buyer beware. Any investor purchasing securities backed by a pool of loans should make sure that the loans really did meet the underwriting standards advertised.