3 Good Ideas for Mortgage Securitization Reform

The last substantive piece of the Treasury's Conference on the Future of Housing Finance consisted of breakout sessions. In each of these discussions a different topic was explored a little bit more deeply by the bankers, regulators, and experts, as they ate lunch. Meanwhile the reporters listened, scribbled furiously on their notepads, and sighed wearily as their stomachs growled from hunger. I attended a session called, "Funding Mortgages and the Role of Securitization." It was a surprisingly productive discussion where a number of good ideas were suggested. Here are three of the best.

But first, why does this topic matter so much? Prior to the housing bubble's pop, the securitization market was an essential source of funding for mortgage issuers. While mortgage bonds backed by the government are still being purchased by investors, the so-called "private label" securitization market -- utilizing pools of mortgages not backed by the government -- has been virtually closed since that time. For the U.S. housing market to function properly again, securitization must return. Yet, investors are still very wary about buying complex bonds backed by mortgages. This is the problem the breakout session sought to address.

A Mortgage Market Infrastructure

One potential solution cited by some to fixing the mortgage securitization market is better transparency and information. If the investors had all of the information they needed to fully analyze the mortgages behind the bonds, then they would be able to better evaluate the securities without having to rely on rating agencies. But one of the session participants said it wasn't that easy. Investors no longer trust the data they're provided. Since fraud was a significant problem during the housing boom, you can imagine why. Now no amount of data will help, as investors worry about its accuracy.

One of the other participants, Federal Reserve economist Diana Hancock, suggested a good solution. The mortgage market needs an infrastructure. Imagine if a broad database existed providing borrower statistics for each mortgage. It could provide information like FICO scores, other outstanding debt, income, etc. The database could also indicate if the information has been verified by independent third-party due diligence. This would not only help confidence in the data, but a robust framework would help investors to better understand how the assets might perform.

Risk Retention for Lenders -- Not Securitizers

One of the ideas championed by many policymakers -- including Congress and FDIC -- is the concept that banks should have some "skin in the game" if they sell assets through securitization. Consequently, Dodd-Frank financial regulation bill requires securitizers to retain a 5% vertical slice of any new mortgage-backed security created.

Some argue that having skin in the game wouldn't have helped prevent the housing bubble, but even if you think it would have, it depends on whose skin we're talking about. The purpose of forcing a bank to retain some of a loan is to ensure that the asset's risk is relatively low. It would be forced to eat some of its own cooking. Otherwise, if the bank can sell an asset entirely, then it doesn't care how risky the loan turns out to be. Yet, if this regulation is aimed at the securitizer, it doesn't really address this problem. That bank may have just been a middleman in the process. Instead, it should apply to the lender that originates the mortgage (or other loan) in the first place. After all, that's where a loan is created -- so that's where excessive risk should be prevented in the first place.

Unified Capital Standards

Fed economist Diana Hancock had another very important suggestion. One of the major problems with the GSEs stemmed from their luxury of having unusually low capital requirements. This was bad for a few reasons. First, obviously those ultra-low capital requirements turned out to be pretty poorly conceived, as Fannie and Freddie quickly burnt through their capital when the market began to falter. Second, lower capital requirements provided the GSEs a competitive advantage over others in the market. They made it even harder for the private sector to originate mortgages as abundantly and cheaply as the agencies.

Hancock's solution quickly solves these two problems. Higher capital requirements (and implicitly lower leverage) would make any government agency involved in backing mortgages more stable. It would also allow the private sector to better compete. Ideally, the private sector would take over entirely, so we should seek to lower any barriers that stand in its way, like differing capital requirements.