Every advance country in the world has some regulation and government involvement in their mortgage market. In 2011 we are going to discuss the proper role of the government and the United States’ mortgage market as GSE reform comes into the public.
James Kwak has two posts, which brought some helpful comments from Arnold Kling. Yves Smith and Zach Carter shoot a round off at Center for American Progress’ GSE proposal. Before this wanders too far into the weeds, there’s a few simple questions that will guide where reform needs to go. Raj Date from Cambridge Winter came up with a decision tree about GSE reform that I think was very helpful. You can zoom into the full tree here:
But I want to focus on the last three options, the options that are most in play:
With that in mind, we can break this into a few questions.
Will There Be Long-Term Fixed Rate Mortgages without the Government?
The first question is whether or not the private market can provide capital for long duration, fixed-rate mortgages. This isn’t a question of whether or not the rate will go up – it will – but whether or not it would be workable at the $10+ trillion dollar range. If so, then with some standardization and minor regulatory tweaking the government can largely get out of the way. This view can be associated with AEI’s Alex Pollock and here is his proposal: Reform of the Housing Finance System.
People are arguing that this can be met without government guarantees because there is significant consumer demand for this product. I’d really like to see more data on this. My instinct is that the prime market would just become subprime writ large, with what we saw over the past decade of non-GSE mortgages. The characteristics of subprime mortgages were that they were de facto bullet loans, designed to be refinanced every few years, not practically callable due to expensive prepayment penalties. From Making Sense of the Subprime Crisis by Kristopher S. Gerardi, Andreas Lehnert, Shane M. Sherlund, and Paul S. Willen, within 30 months over 80% of subprime loans have refinanced. That transfers extensive liquidity risk to the household, and between transaction costs and penalties these aren’t great for building equity.
Also having a fixed-rate duration of over 20 years isn’t very common internationally. This is from The American Mortgage in Historical and International Context by Richard K. Green and Susan M. Wachter, as of 2005:
But I could be convinced otherwise.
Should We Care?
The second question is whether or not it matters. Should we care if people can get a 30 year fixed rate mortgages? Center for American Progress’ David Min argues yes here. The argument goes:
– First, the 30-year fixed-rate mortgage provides cost certainty to borrowers, which means they default far less on these loans than for other products, particularly during periods of high interest rate volatility.
– Second, the 30-year fixed-rate mortgage leads to greater stability in the financial markets because it places the interest rate risk with more sophisticated financial institutions and investors who can plan for and hedge against interest rate fluctuations, rather than with unsophisticated households who have no such capacity to deal with this risk and who are already saddled with an enormous amount of financial burden and economic uncertainty.
– Third, the 30-year fixed-rate mortgage leads to greater stability in the economy because short-term mortgages are much more sensitive to interest rate fluctuations and thus much more likely to trigger a bubble-bust cycle in the housing markets. Indeed, there may be reason to believe that a primary cause of the recent housing bubble-and-bust cycle was the rapid growth of short-duration mortgages during the 2000s, which caused U.S. home prices to become more sensitive to the low interest rate environment created by Alan Greenspan’s Federal Reserve.
It’s a good piece worth your time if you want to learn the arguments. Mortgages like these pool certain risks, moving them to those most optimal for handling them. It decreases the likelihood of boom-bust cycles in the housing markets, and it makes monetary policy more effective.
If There’s a Role for the Government, What Does It Look Like?
A lender consists of three people: the person who provides the funds, the person who takes the interest rate risk, and the person who takes the credit risk that they won’t be paid back. Usually this is the same person. Derivatives can split these off, and the government can absorb these risks as well.
In the lead-in to the crisis, most people thought that the GSEs would blow up because of this interest rate risk. Instead it blew up because of credit risk. If you look at reform packages for the GSEs around 2004, you’d see reforming the interest rate risk as key. Not so anymore. See this excellent post by John Hempton for more.
If we decide that 30 year fixed rate mortgages are worth subsidizing, what is the way to go about it? The Center for American Progress proposal mimics the structure of FDIC: regulated firms, CMIs in their language, bid the price of credit risk, which is backstopped by the United States government. Losses are covered by fees assessed to the firm, and in the case of a CMI failure, remaining firms. This is similar to how FDIC handle’s its backing. From their proposal:
How you think of this proposal will depend on how you think of these firms vis-a-vie the recent crisis. We have Too Big To Fail firms that had to have emergency buffers put behind, and we have neighborhood banks that are being taken down just fine, with the cost of backing checking accounts coming out of the emergency fund which will be paid back by other firms. There’s also issues of regulatory capture and issues related to CMIs in practice. Will they be properly regulated? How will they be resolved in cases of failure?
Another approach, one Raj Date developed, is to subsidize the interest rate risk but leave the credit risk off entirely to the markets. Raj Date presenting his approach to GSE reform at the Roosevelt Institute’s Make Markets Be Markets conference, along with his recent Starting Over brief on GSE reform. The general idea is to make the interest rate swap much more transparent, seed a covered bond market to begin a transition to a post-crisis mortgage landscape if necessary, use explicit cash transfers for housing policy and leave the private market to do the rest. But does this just create an implicit guarantee by saying there’s no guarantee? Is it better to just formalize the relationship, like with FDIC and checking? Here’s how to think of what the GSEs have done, and would do under this approach:
That’s the landscape, with the consensus forming around a CAP-style program.