House under construction

Housing Perspectives

Research, trends, and perspective from the Harvard Joint Center for Housing Studies

What the FHFA Needs to Get Right in Its GSE Capital Rule

In July 2018, the Federal Housing Finance Agency (FHFA), regulator of the two government-sponsored enterprises (GSEs) Freddie Mac and Fannie Mae, published a proposed rule of the minimum capital the companies would need, to apply when they exit conservatorship at some point in the future. After seeking feedback from the public, the FHFA began analyzing those comments in November 2018.

The rule still has not been finalized.

And yet, the capital rule is, as much as anything, the linchpin to moving forward on housing finance reform in the US. The recently-released Treasury “Housing Reform Plan” – or any other plan – cannot reasonably move forward, past a certain modest point, without the capital rule being finalized. After all, if you don’t know how much capital is needed to escape conservatorship, how effectively can you plan the escape? My new paper, “Four Big Things the FHFA Needs to Get Right in Its GSE Capital Rule and Why It Will Take Political Courage to Do So” gets to the heart of the issues that must be properly addressed if the capital rule is not just to be finalized, but to represent the best of professional regulation without politics biasing the result into over-capitalization (which will unduly raise the cost of mortgage credit) or under-capitalization (which puts the financial system at risk).

In fact, the FHFA’s proposed capital rule will be the first set of such requirements developed since those put in place after the 2008 financial crisis have had some time to run. That means the rule can build upon what has been learned in the decade-plus since. Also, for better or worse, the FHFA proposed capital rule also builds upon a long history of financial institution capital regulation, both in the regular banking system and the specialty mortgage system (to which the two GSEs belong), which have had both successes and failures since the era of modern capital regulation began in the 1980s.

And in the specialty mortgage system, there have arguably been more failures than successes.

Against, this backdrop, the FHFA will need to courageously address four tough issues that appear to me to have been politicized more than is desirable:

Tough Decision #1:

Should the capital rules be the same as they are for Systemically Important Financial Institutions (SIFIs), all of which are banks, or just be consistent – i.e. the same capital for the same underlying risk, recognizing that the GSEs are not banks?


Since the GSEs have, like insurance companies, a different business model and risk profile than banks, it makes no sense to directly apply bank-centric SIFI capital formulae to them. What does make sense is to use the Federal Reserve’s bank-centric formulae as a starting point from which to get to a goal: a GSE should have the same amount of capital as a SIFI bank for the same amount of underlying risk. In particular, the GSEs should be able to pass a SIFI stress test: capital sufficient to absorb a severe adverse economic scenario (in this case, house prices down by 25%, among other things) plus have a regulator-determined “going-concern buffer” to retain market confidence.

Tough Decision #2:

Whether to propose leverage ratios (which by definition are not directly risk-related) that are designed only as back-ups in line with the Basel agreement, or the higher ratios which, as an international outlier, apply to US banks today, and then whether to apply them via a traditional simple ratio, or choose the bifurcated (i.e. two ratio) approach also proposed?


The FHFA will need the courage to adopt anything with the specific ratios proposed, as all the numbers seem so low in comparison to those used for banks. But as they do treat the leverage ratio the proper way – as a back-up and not a primary (and potentially binding) driver of the level of capital – it is the right thing to do. It then will need additional courage to adopt the bifurcated approach, which reflects the specific business model of the GSEs and has much less risk of creating major economic distortions, especially by potentially rendering credit risk transfer uneconomic. Given the strong tendency I have seen among politicians and pundits to talk about the GSEs as if they were banks and to look simply at the accounting-driven simple leverage ratio, as opposed to the more complex risk-based approaches, the reaction to these choices could be strongly negative.

Tough Decision #3:

How should the capital system employ the loan-to-value (LTV) ratio, which is particularly important in measuring the risk on a portfolio of mortgage loans: unchanging over time (OLTV; always using the “original” value of the home when the mortgage was purchased); or changing over time (CLTV; reflecting how house values move up and down) and which produces pronounced procyclicality in capital requirements?


For a number of reasons outlined in the paper, OLTV is a very poor choice for measuring credit risk on GSE mortgage guarantees over extended periods of time. Using CLTV is not only the economically accurate way to measure risk and required capital at all times, but also the only one likely to maintain market confidence in a stressed economic environment. In fact, the test of the quality of a capital system is found not in good times – it is in the worst times that one finds out if the system was well designed or not!  The FHFA must therefore address head-on how to deal with the strong procyclicality of the business, accurately reflected by using CLTV.

Tough Decision #4:

How to deal with the pronounced procyclicality of being a mortgage guarantor in terms of its impact on capital requirements?


As there is no well-established approach in the banking system to deal with this type of pronounced procyclicality, I recommend one that is particular to the specific business model of the GSEs. It is based upon laying off the procyclicality risk on the marketplace via modifying how credit risk transfer transactions are structured. This also means the cost of addressing the pronounced pro-cyclicality is very modest, as the market will price in how remote are the odds of a major stress event actually occurring.

In the paper I explain and elaborate on the answers to the above four questions. Unfortunately, given the long tradition of politicization of the mortgage system in general, and Freddie Mac and Fannie Mae in particular, it will require real political courage on the part of their regulator to do this properly. But it is important for the FHFA to have that courage, as so much is riding on it, including its own reputation as a professional regulator, not swaying with the political winds, as so many of its predecessors have.