House under construction

Housing Perspectives

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

Why Is the Administration Not Talking About Utility-Style Regulation of G-Fees?

The latest administration pronouncements about the reform of the two giant government-sponsored enterprises (GSEs), Freddie Mac and Fannie Mae, as delivered in the President’s Memorandum on Federal Housing Finance Reform, issued in March, and also in public and private comments by Treasury and Federal Housing Finance Agency (FHFA) officials, have mentioned many principles about how GSE reform should work.

Prominent among those principles is an emphasis on competition, of trying to somehow have more than two GSEs. A separate set of principles from the Chair of the Senate Banking Committee, Senator Michael Crapo (R-Idaho), with whom the administration states it wishes to work, even had a deadline of sorts for more competitors to be created. Treasury officials, recognizing the King Canute nature of this (that’s the story where the king, to counter excessive flattery from members of his court, demonstrates how he could not stop the tides from coming in), talked in some meetings I attended of setting up mechanisms to allow additional firms to enter as easily as possible, but of course knew it could not be forced or guaranteed. Mark Calabria, the new head of the FHFA (the regulator of Freddie Mac and Fannie Mae, along with the Federal Home Loan Banks), also talks about competition and having more than two GSEs in public speeches, albeit at an abstract level.

The industry, watching all this, has been uncertain how to react, as it knows that the barriers to entry to compete with $2 trillion Freddie Mac and even larger $3 trillion Fannie Mae are high, possibly even insurmountable. (I could easily argue they are indeed insurmountable, but one never quite knows, as industries and technology evolve, so I personally consider the barriers to be very high; I will explain these barriers and why they are so difficult in a future post.) Even if they could eventually be surmounted, I have not encountered any knowledgeable people who think it could happen quickly.

So, the government should clearly plan for the continuation of the two GSEs, at least for several years, if not forever. But if there are only two, would there be effective price competition when it came to the setting of single-family mortgage g-fees? For those unfamiliar, g-fees are guarantee fees, and represent the charge to lenders, above what capital markets investors charge to finance mortgages, to reflect that the GSEs guarantee the credit risk on the mortgages to those investors. They are at the core of what the two GSEs do and generate their primary source of revenues and profits.

I think the answer is: likely not. Having just two competitors in such a situation is ripe for the evolution of implicit collusion (also known as tacit collusion). Implicit collusion in price setting is well known in US economic history. For example, in the decades after World War II, many industries developed a price leader (usually the largest company in the industry). The leader would raise prices from time to time (often after a labor union contract negotiation), and others would match it, trusting the leader to set a price that would be good for all. There were no meetings to plot this, no secret messages being passed; it just developed over time as the mutually advantageous and quasi-oligopolistic behavior among firms in an industry. And the two GSEs, if privatized again, would be ripe to engage in just such price-setting implicit collusion, with the possible risk of g-fees being materially higher than they need to be.

Now, you might ask: before the companies went into conservatorship, did they implicitly collude? There is some general belief in the industry that they did, although seemingly less to set g-fees at high levels (they were very low, compared to where they are today) and more to support a market split (averaging 38 to 40 percent for Freddie Mac, the rest for Fannie Mae), which allowed them to be… well…  lazy competitors. What I have found is an almost universal belief that g-fees were set, at that time, more on a political basis than an economic one (i.e. to be as low as possible), and that the companies fed their need for a rising stock price via increased revenues and profits not from g-fees but from their subsidized investment portfolios, which became astoundingly large and ever-growing.

For those not familiar with one of the largest unintended financial consequences ever, the GSEs, when they were privatized by the US government – with the implicit guarantee by the government to support the issuance of mortgage-backed securities at low rates of interest – were also able to issue unsecured obligations with that same support. And thus, they borrowed, without limitation, almost as if each were the US Treasury. This made the borrowings so cheap that the two companies were able to buy mortgage securities (including their own!) and make a large profit on the interest rate spread between the asset yield and their cost of funding those assets. It was so large, in fact, that the investment portfolio became the biggest source of profit for each of the two GSEs, and even more so at Freddie Mac than Fannie Mae. (I was able to verify upon becoming CEO of Freddie Mac that the source of the investment portfolio profits was substantially due to the cheapness of the funding, and not brilliant investment strategy.)

As a result, critiques of the industry, with some accuracy, sometimes talked about the GSEs as subsidized hedge funds with mortgage guarantee businesses attached. At the height, in 2006, the investment portfolios of the two GSEs amounted to about $1.5 trillion, which was almost double the balance sheet of the entire Federal Reserve at the time. This was an unprecedented size.

So, g-fees were kept very low back then, which satisfied politicians and industry representatives who wanted (and still do) cheap credit for homeowners, and also reflected the generally understated (compared to today) level of capital required for financial institution risk-taking. Profit growth came instead from the subsidized investment portfolios. That growth was so significant that the GSEs even advertised themselves as growth stocks to the investment community.

But now, after a decade in conservatorship, this “deal” has changed. The GSEs were required by the terms of the 2008 government rescue to shrink their investment portfolios dramatically, over time, and to become solely supportive of the guarantee business, with no more discretionary investing. The size of the portfolios is therefore down by roughly 75 percent since 2006, to just about 10 percent of total GSE assets; this is a far cry from when it got to over 30 percent pre-conservatorship.

So, the companies are now, as they always should have been, focused on their core mission of operating a mortgage guarantee business, which in turn generates most of their profits. That means, if or when conservatorship ends, they will be very focused on the g-fee to produce the profits that support their stock price, just like any other public company.

And with just two GSEs, and one the clearly larger firm, it is almost a classic set-up for implicit collusion in price setting to occur, resulting in higher g-fees than needed. (The implicit collusion could also manifest itself in price cutting to deter new competitors if any truly looked to be successfully entering the business.)

So, what is to be done?

In the last several years, some observers of GSE reform have suggested treating them like utilities, such as a local electric or water utility. Those types of firms are generally subject to rate-of-return regulation by the local state public service commission, which sets prices (called rates in this context) to the consumer to generate, after a reasonable allowance for expenses, a proper return on the capital employed (usually called a “fair return” in legislation). If the approved rates are too low, the utility will find itself unable to finance the new capital investments needed to provide reliable electric power or water service, which causes problems for the public service commission when the delivery of the regulated service deteriorates. If the rates are too high, public reaction will make things very difficult for the public service commission, politically. It seems to generally balance out, and has worked adequately (if not perfectly) throughout the United States for about a century now. To me, that classifies as tried and true!

But what is not well known, a classic demonstration of the “knowledge gap” between conservatorship insiders and everyone else, is that the FHFA has developed the core of just such a utility-style rate-of-return regulation of the g-fees of the two GSEs. It took multiple years and required detailed work on defining g-fee revenues (which can sometimes be commingled with other revenues) and establishing the capital employed against them. But it was in place by 2018 and refined even more for 2019. And it was, when I left Freddie Mac at the end of June, working reasonably well. It was in the form of a floor for the returns earned on guarantees (in conservatorship, the tendency is to cut g-fees to win market share), but could easily be, post-conservatorship, in the form of a target or ceiling (where the tendency would be to raise g-fees to increase profits and the stock price).

Today we are waiting for the administration’s plan on how to exit conservatorship, which is promised “soon.” The consensus expectation is that it will not be a fully detailed plan, but more likely a comprehensive list of chosen design principles, with specifics on some portion of them.  But, based on all the comments by officials from the administration and the FHFA, there is no focus on utility-style regulation of g-fees.

In my view, however, given how hard it will be for competition for the GSEs to develop, especially in the near term, there should be such a focus. It can always be dropped later if there is an evolution where two or three additional GSEs enter the business successfully. But until then, let’s adopt the tried-and-true utility regulation approach to setting g-fees to avoid unfortunate and undesirable implicit collusion, the unintended consequence of which could be unnecessarily high g-fees.

And since the FHFA already has a working utility regulation approach in place, it should be an easy lift to do just that.