The Future of the GSEs

Paulson recently spoke on the subject of his actions regarding GSEs and the his recommendations for their future. Given that he covered two of my favorite subjects (housing and P3s), it got my attention.

Paulson makes 3 broad claims in this article that I disagree with, one of which is worth emphasizing:

1) it is the GSEs' mixed-incentive structure that allowed them to become so big. This is an odd place for Paulson to stake his claim given that he turned the TARP into a mixed-incentive structure...

2) the GSEs can help the economy return to normalcy by keeping mortgage rates low and by mitigating foreclosures. Artifically low rates were part of what got us where we are, but addicts will be addicts...

3) the question re: the GSEs' future is how much we want the government to reduce mortgage interest rates by taking on mortgage credit risks. It is this issue (i.e. the role of GSEs in housing policy) that I want to address.

Yes, one way GSEs can affect housing policy is by taking on mortgage credit risk. Whether or not this credit risk is appropriately priced and/or the benefits of government-sponsored funding are passed through to borrowers is a question that has been long debated. Yes, it is important to ask whether the private sector couldn't address most issues of credit risk adequately without GSEs. But no, I don't think credit risk is the only, or even primary, role for the GSEs.

The GSEs' primary role is actually contained in Paulson's reference to covered bonds as a potential solution to the current situation. Without government involvement, most mortgage rates and mortgage terms would be variable (rather than fixed) and short-term (less than 30 years). Why? Because this would match the rate and term obtained on funds used by financial institutions and mortgage market participants to create mortgages.

Are variable rates and short maturities good public policy? For households that can bear financial risk, those kinds of terms are fine and perhaps even preferable. But for households on fixed incomes, limited incomes, or purchasing their first home and needing stability in their housing cost, short-term variable rate debt can be deadly. And it is here, in the world of creating more long-term fixed rate mortgages than would otherwise be generated by the market, that GSEs and housing policy can come together.

It is generally misunderstood that GSEs borrow on variable rates with relatively short maturities and use it to fund fixed rate long maturity mortgages (they also used it to fund a lot of short-term variable rate mortgages as well but that is a different issue). That's why we all care that debt holders might no longer want to buy GSE debt - if the GSEs financed all debt with rates and terms matching the mortgages that were purchased, no one would care that the demand for GSE debt might decline. But, when the GSE debt has a short maturity that is due and subject today's rates and the mortgage asset has a long maturity at a fixed rate, well, Houston...we have a problem and it looks a lot like the last one (S&Ls in the 1980s).

Why would any entity create a mismatch? Because it generally is financially profitable to do so as long as short-term rates remain relatively low, long-term rates are relatively high, and there is not a liquidity problem with rolling the short-term debt over each period (i.e. the achilles heal in the current environment).

This is the real question for the next administration and financial markets to resolve - does changing who creates and uses mismatched funding (i.e. S&Ls, investment banks, commercial banks, the private sector, or GSEs) make a difference in who carries the risk, how well it is handled, and the impact it has on the market place? I am going to suggest that it possibly does, but in a sort of indirect way. What are long-term rates comprised of: inflation expectations, risk premiums (i.e. stability of returns), overall level of rates, etc. No one entity controls those factors, but who has the most influence on them? Monetary and fiscal policymakers. So, to the extent that mismatched funding is best influenced by monetary and fiscal policymakers, it seems that is also best used by monetary and fiscal policymakers. In other words, why should the GSEs or the government take on mortgage credit risk or issue adjustable rate loans that are less than fully amortizing? The private sector can do that just fine. But, it is also very unlikely the private market will address the power and stability of a fully amortizing 30-year fixed rate loan to the level desired by public policy.

Underneath the surface of Paulson's political argument (i.e. this problem started before I came, I tried to fix it early on, I couldn't get it done because of politics, that's why I took the extraordinary steps that I did) lies the political-economy argument that should be addressed by this Congress and this administration - mismatched funding is a powerful but dangerous game that thrives on imbalances in the financial markets. Understanding that point is critical for determining the best way forward for the future of the GSEs and housing policy.