Rethinking The Mortgage Industry: Is There A Better Way To Subsidize?

Wharton’s Tim Landvoigt discusses how underpriced government mortgage guarantees contributed to the housing crisis.

American homebuyers were riding high in the housing boom of the early 2000s, when cheap, easy credit let just about everyone obtain a mortgage. But when the housing bubble burst, millions of Americans found themselves upside down in their mortgages. Indeed, the housing crisis of the last decade caused tremendous damage for cash-poor homeowners who needed years to get back on solid financial ground. It also left behind a trove of data for researchers to sort through. Wharton finance professor Tim Landvoigt is one of those researchers whose analysis could help prevent similar crises. He spoke to Knowledge@Wharton about his research and what he’s learned.

net worth sweep Fannie Mae Freddie Mac
By User:AgnosticPreachersKid (Own work) [CC BY-SA 3.0], via Wikimedia Commons
Knowledge@Wharton: Tell us about the focus areas of your research?

Tim Landvoigt: One of the areas that I do research in is housing and mortgage finance. I have several papers that are concerned with the question: What were the main drivers of the housing boom in the early 2000s? The candidate explanations that people have come up with are cheap credit, easy access to credit and high house price expectations — expectations about large future gains in house prices.

Broadly speaking, I have several papers that come to the conclusion that cheap credit and easy access to credit are enough to generate a housing boom of this size. You don’t need exuberant expectations about future house prices. One of the main insights empirically of these papers is that we saw a larger housing boom, a larger price gain in the low-quality segment of the housing market. Initially, relatively cheap houses experienced much larger price gains than the higher end of the market. That is consistent with the theory that the relaxation of credit constraints and cheap credit were the primary drivers of housing demand for relatively less-wealthy people, who were then the natural buyers of these relatively cheaper houses. The housing boom was much smaller in the higher-quality segments of the market.

“Cheap credit and easy access to credit are enough to generate a housing boom of this size.”

Knowledge@Wharton: One of the questions that has come out of that housing crisis has been what to do next with Fannie Mae and Freddie Mac, and you have a paper that looks at that. Can you talk about your key takeaways?

Landvoigt: That paper is called “Phasing Out the GSEs,” which is a somewhat provocative title. [GSE stands for government-sponsored enterprise.] But what we’re really looking at in that paper is whether we should increase the guarantee fee that these government sponsored enterprises charge banks when they sell the loans in the secondary market. Our conclusion is that the guarantee fee has been much too low historically. It’s basically a very large subsidy to mortgages originated by banks that then sold to Fannie Mae and Freddie Mac. We show theoretically that that would provide incentives for banks to take on more risk in other areas of that balance sheet. Because we’re removing a lot of risk from the mortgage part, the banks search for that risk somewhere else.

The conclusion is that we should increase those GSE fees to the extent that they are a fair insurance fee for the risk in these mortgages. To some extent, the GSEs have already adopted such policies.

Knowledge@Wharton: What are the practical implications from your paper? What could we do to make sure another housing crisis doesn’t happen?

Landvoigt: To the extent that they have already increased the overall level of the GSE fee, and they are charging these guarantee fees more conditional on the true riskiness of the loans, those are already steps in the right direction. But the paper goes further and says that it might not be ideal for the government to guarantee the risk of these mortgages. If the government wants to subsidize housing or housing affordability, it should do it in some other more direct way, and not through this kind of indirect way that messes with the incentives for banks to take on too much risk. I think the overall takeaway is that if you want to subsidize housing, don’t do it through a guarantee fee for mortgage originations. Just increase the mortgage interest rate deduction or do something more salient and direct.

Knowledge@Wharton: Where you are going next with your research?

“If you want to subsidize housing, don’t do it through a guarantee fee for mortgage originations.”

Landvoigt: Another interesting project that is related to this whole area of housing and mortgages that I’m working on is about shared appreciation mortgages or shared home equity mortgages. About 80% of all mortgages in the U.S. are fixed-rate mortgages, either 30-year or 15-year fixed-rate mortgages. These mortgages are debt contracts, so you have to repay a fixed principal amount that amortizes over 15 or 30 years. That means if the value of your house drops a lot, like it did during the crisis, then you might have negative home equity, you might be underwater, and that gives you incentives to default.

An alternative mortgage contract might be one where the bank explicitly takes on some of this mortgage risk. If your house price drops a lot, then the bank will automatically mark down the mortgage principal that you owe. But if your house appreciates a lot, then the bank will get a share of that. We shift some of this risk to the bank, which will then avoid foreclosures and defaults and crises. But it will also make the mortgage more expensive because the bank will have to charge the homeowners for the risk. So, there are several considerations here. On the one hand, it might be good because we are avoiding these foreclosures and defaults. But on the other hand, we’re now saddling the banking system explicitly with this risk.

If banks are already kind of capital constrained or worried about having too much risk on their balance sheet, they might not want to make these loans, or they might become too expensive. We’re trying to figure out, again in a large, calibrated macro-model, which of these effects dominates. Is it good or bad overall to offer these contracts in the market?

Article by Knowledge@Wharton