When Will Government Control of Fannie Mae and Freddie Mac End? (Part 2)

Research & Policy | July 25th 2022 | Donald H. Layton

Glendale, AZ aerial view

Part 2 - Many Years Still to Go; Steps to Move Towards the Off-Ramp

The two government-sponsored enterprises (GSEs), Freddie Mac and Fannie Mae, were put under government control in September 2008, at the height of the financial crisis, when the market was losing confidence that they could service their $5 trillion-plus of outstanding debt.  This government control was implemented by putting the two companies into “conservatorship”, a legal status whereby their regulator, the Federal Housing Finance Agency (FHFA), became able to fully direct each company’s affairs by stepping into the shoes of its shareholders and board of directors. Additionally, the Treasury restored market confidence in the two companies by entering into a Preferred Stock Purchase Agreement (PSPA) with each, legally committing to support their financial strength.

As described in Part 1, when the GSEs were put into conservatorship under the Bush administration, the prevailing view was that fundamental reform of their operations was a necessary prerequisite to ending government control. Since only Congress could implement such changes, conservatorship lasting only briefly seemed unlikely. The Obama and Trump administrations then made decisions that kept pushing the timeframe further into the future, with the notable exception of the latter authorizing the two companies to build capital by retaining earnings starting in September 2019. The Biden administration has thus far not publicly taken any additional steps to head towards the off-ramp from conservatorship, although importantly it has not stopped the retention of earnings.

In just two months, it will be the 14th anniversary of that government control, with no end in sight. Part 2 addresses two related issues.

  • First, how much longer can we expect government control to last? While it is impossible to predict when the FHFA and Treasury, the two organizations that have to agree to commence and then implement an exit, will embark upon such a path, there are certain realities of how the equity markets work which will put a multi-year floor on the time required for implementation.
  • Second, what are the most important steps that can be taken in the near term to prepare the companies for such an exit to happen as expeditiously as possible, once the FHFA and Treasury decide to go ahead?

With respect to the second issue, it is notable that Sandra Thompson, the newly-confirmed Director of the FHFA, has publicly indicated she is interested in getting the two companies prepared for an eventual exit from government control. Her first concrete action in this direction was to amend the minimum required capital rule - issued by the FHFA when Mark Calabria was its director - to eliminate its bias against credit risk transfer (CRT) transactions. This will result in more such transactions being completed and thus reduce the future need for capital, working to shorten the off-ramp to ending conservatorship (as well as reducing taxpayer exposure to GSE losses in the meantime).

The Market Realities of Raising Equity Capital

After all the delays described in Part 1, if and when the FHFA and Treasury eventually do decide to have the GSEs (presumably retaining all the reforms made since 2008) exit conservatorship1, it will be necessary to consider how to actually implement such an exit. The one absolute implementation condition is that the two GSEs will need to have more capital than the minimum regulatory requirement. While many housing finance policy people talk about this in the abstract, actually doing it will face some inconvenient realities of the equity market - one of which is that raising the necessary capital will almost certainly require many years to accomplish. Consider the following:

  1. The amount required is immense, far beyond anything ever seen before. As stated in Part 1, my rough estimate of the amount required is $150 billion for the two GSEs combined.2 For context, that is roughly eight times the $17.9 billion size of the largest IPO in history for an American company.3 And the actual amount of capital required could of course be much higher than my estimate. Fortuitously, GSE profits have been strong through the pandemic (in particular reflecting rapidly rising house prices), so that their net worth, since earnings were allowed to be retained starting September 2019, is now a combined $83.5 billion4, already more than halfway to my $150 billion estimate. This validates the wisdom of the 2019 PSPA amendment to allow earnings retention.  
  2. Equity market investors have basic requirements that cannot be met during conservatorship. One of them is that they have the usual shareholder rights, the most core of which is the ability to vote on who serves on the Board of Directors to direct the affairs of each GSE. Unfortunately, in conservatorship, investors have no such rights; instead all the rights of the shareholders belong to the FHFA as conservator. Therefore, institutional investors will have no interest in purchasing shares through a new issuance of equity - and especially a mammoth-sized one - while the companies are in conservatorship. This means that a strategy to capitalize the two GSEs must start first with a long period of retaining earnings, fully (or nearly so) capitalizing the companies via this route. With today’s existing net worth of $83.5 billion, and using some back-of-the-envelope estimates of normalized annual earnings, this means it could easily be another four to five years to raise enough equity.5
  3. The cost of equity raised must be as low as possible. The cost of equity is critical for holding down the level of the GSE guarantee fee (G-fee), which is one of the primary components of the interest rate on mortgages. The G-fees charged by the GSEs have three cost components: (1) the operating expenses of the GSEs, (2) the provision for loan losses that are required to be set aside under accounting rules, and (3) the “cost of capital”6 on the capital supporting the risk of the guarantees. That cost of capital is the largest of the three. From a public policy perspective, it is therefore imperative that the cost of capital for the GSEs be as low as that enjoyed by the largest, most liquid and well-traded financial institutions (after adjusting for their relative riskiness) in order to avoid G-fees being higher than necessary. Although not well known, current G-fee levels are based upon just such a “best-lowest cost” concept.7 But if there is a strategy to shorten the off-ramp from conservatorship by pushing equity issuances when there is limited market appetite for the shares, then the share price will need to be reduced to attract a sufficient number of investors. The result is that the cost of capital would be higher, and potentially much higher, leading to increased G-fees. Given the public policy objective of as low a G-fee as reasonably possible, this means there is no consequence-free short-cut to raising the equity quickly by pressing the public markets to buy GSE shares at some discount.

The result of these realities is that recapitalizing will most likely occur through retained earnings over the course of the next four to five years, putting the GSEs on track to begin winding down conservatorship no earlier than about 2026. If more than $150 billion is required, or if there are other significant steps required outside of recapitalization, the timeframe could be pushed out further. This conclusion has been unwelcomed by many people involved in GSE matters who wanted it to happen far more quickly.8 But so far, it has held up rather well.

Surprisingly, Government Control Will Outlive Conservatorship.

After the GSEs are fully recapitalized and various other requirements are met, they can be released from conservatorship, according to FHFA regulation. It is commonly assumed that this would be the end of government control. But the reality is more nuanced. Instead, the location and legal mechanism of government control just changes: instead of the FHFA as conservator controlling the two companies, Treasury will take the reigns as their majority shareholder.

This stems from the fact that Treasury, as one source of compensation for providing its financial support to the companies via the PSPA, was awarded warrants on the common equity for 79.9% of the shares of each GSE. This means that Treasury, upon the exercise of the warrants when conservatorship ends9, will have 79.9% voting control of the GSEs. In addition, Treasury will then also own over $200 billion of senior preferred shares in the two companies10. There is uncertainty as to how Treasury would deal with the senior preferred equity ownership position (an explanation of which goes beyond the scope of this post) but the only precedent is for Treasury to convert the position to additional common shares, and thus end up actually owning well more than 79.9% of the shares, with the associated stockholder voting power that goes with it.11.

To actually fully wrap up and conclude the rescue of the two GSEs, Treasury will then need to sell its shares to the public marketplace. (This is called a “secondary” offering, as it is one shareholder selling to others; it does not directly impact the actual balance sheet of the issuer of the shares.) As described above, the size of the positions to be sold by Treasury are so huge versus even the largest IPO ever done that this will not take place overnight. A reasonable minimum estimate, which is highly dependent upon equity market and mortgage market conditions at the time, is at least three years - but potentially much longer.12

Somewhere along the path of Treasury selling its shares, it would cease to be the controlling shareholder.13 That is when, truly and finally, government control would end. Given that we’re already 14 years into this journey, it does not seem unrealistic to expect continued government control until at least 203014, and possibly even much later.

Recommendations to Finally Get onto the Off-Ramp

Given the potential for these extended timeframes, it behooves everyone involved to do everything that can be done ahead of time, to avoid unnecessarily extending the life of government control even further. FHFA Director Thompson has stated that this is among her priorities. Here are, in my view, the four most important actions to take:

  1. Continue to build capital by retaining earnings. The best and most important decision made to help the GSEs be financially stronger, more safe-and-sound, and also closer to exiting government control was to have them retain their earnings to build capital beginning in September 2019. The Biden administration has wisely continued this policy. It is, in my view, the single most important priority for the FHFA (and also Treasury) to retain it as a permanent feature of the conservatorship until the companies are considered fully capitalized.  
  2. Develop a new and credible minimum required capital regulatory rule. The current minimum capital rule has been widely (and accurately, in my view) criticized as being non-credible, in particular for being far too high. That is most apparent in how it is wholly inconsistent with the results of the official government “stress test” for each GSE, i.e. the capital rule being well too high when compared to the low losses calculated by the stress tests.15 (Upon the release of the upcoming stress tests, which are expected in August, I plan to post specifically about this, and how the stress test results are more credible.) Thus, it is necessary for the FHFA to develop a new rule that is reasonably consistent with the stress test results.16 This will allow more accurate planning and projection around the conservatorship ending, so the sooner it is developed the better. A new rule can also determine what G-fee levels should be during the remainder of conservatorship, a major policy concern given current mortgage rates.
  3. Develop the fee for Treasury support of the GSEs that will be implemented no later than the end of conservatorship. Prior to conservatorship, the GSEs benefitted from the “implied guarantee” of their debts by the US government. The consensus in Washington was and still is that this arrangement was highly undesirable, especially as taxpayers received no compensation for their support. This changed with the advent of the PSPA, which formalized Treasury’s role and used various mechanisms to compensate the taxpayer via payments to Treasury.17 Today there is a strong policy consensus that, post-conservatorship, the two GSEs should pay an overt fee to Treasury for its backing.18 That fee needs to be agreed on by both FHFA and Treasury sooner rather than later, as it impacts some core operating requirements of the GSEs, e.g. what G-fees need to be and possibly how fast retained earnings will accumulate.
  4. Develop the needed FHFA apparatus for implementing the “utility model,” i.e. price regulation, post-conservatorship. As explained in Part 1, the leading model for the “by administrative means” ending of conservatorship - the only one being contemplated today - is based upon the utility model, where the FHFA sets G-fee pricing (and other terms) in a fashion similar to how electric utilities are regulated by the states. Thus, it behooves the FHFA to develop ahead of time all the apparatus needed to be such a price setter. Keying off what state-level electric utility regulation looks like, this means the FHFA has to develop things such as: (1) a method to calculate the “fair” rate of return19; (2) what adjustments, if any, are needed to calculate the capital being employed versus the accounting statement determination of net worth; (3) what expenses if any, might be disallowed20 in calculating what G-fee to charge; (4) how to have public input into the process (at the state level, this is usually via public hearings with testimony); and (5) how often to review and reset the pricing, e.g. annually, or by some other metric. (A full list of all the requirements for a proper utility-style regulation apparatus would be quite long. Fortunately, much of it can be almost wholly copied from the relevant state regulators, who have been doing it for about a century.) The FHFA will also have to develop the legal guardrails around all this so that potential investors in the two GSEs have comfort that the price regulation will not, after they invest their money into the two companies, be politically manipulated at their expense. Adding it all up, it’s easy to imagine constructing this regulatory apparatus could take years, - so the FHFA should get started while the GSEs accumulate capital through retained earnings.21


In conclusion, an exit from government control for the two GSEs is probably in the rough range of 2030, and potentially even later. As described in Part 1, the timeframe of the GSEs exiting government control was repeatedly extended by key policy decisions made by the four presidential administrations since it began in 2008, with the notable exception of permitting the retention of earnings to build capital. Looking forward, as discussed in this post, I have described how actually implementing the end of that government control, once the FHFA and Treasury decide to firmly head for the off-ramp, is going to take many years, mainly due to (1) the time it takes for enough additional earnings to be retained to meet a minimum capital requirement, plus (2) how long it will then take for Treasury to dispose of its control-level ownership stake in the two GSEs.

But behind this is the key assumption that the government actually wants to end its control of the two GSEs. That is, in fact, a questionable assumption as many political and interest groups seem to like today’s status quo. First, they like having the two GSEs, which are stockholder companies chartered by Congress to play a pivotal role in housing finance, firmly under policy and political control through the FHFA. Second, the mortgage industry in particular likes that the GSEs, being structured as stockholder-owned companies, pursue innovation and competitiveness far more aggressively than a typical government agency22. Thus, to many it seems a “best of both worlds”. On top of this, there is considerable fear that the GSEs, given their pre-conservatorship history, might revert, upon exiting government control, to being the highly politicized lobbying powers they were previously.

So, in fact, the GSEs being under government control via FHFA conservatorship might go on as a permanent feature of our housing finance system, even well past 2030, with no return to private ownership being desired or implemented. It is a possibility that cannot be dismissed.


[1] In fact, there is some sentiment that it would be good policy to keep the GSEs in conservatorship permanently. See further discussion in the conclusion of this post.

[2] By contrast the current FHFA-approved capital rule calls for more than double this amount, although it is not regarded by most (based upon the feedback from the public when it was proposed) as a credible calculation. The capital rule was produced under the FHFA directorship of Mark Calabria and seems to reflect, in my view, not mainstream thinking but his preferred policy outcome to force the GSEs to shrink (known as “shrinking their footprint”) as unduly high capital requirements will cause them to lose volume to other providers of mortgage financing.

[3] Visa in March 2008.

[4] As per the March 31, 2022 financial statements of the two GSEs, the latest currently available.

[5] Such an estimate is very vague. It depends upon a lot of assumptions, including how much CRT transactions reduce the need for capital, when the government institutes a fee for its support going forward, and what the level of that fee might be, etc. And this is before all the usual uncertainties associated with predicting profits out more than a few quarters. It also does not reflect that the GSEs need a cushion above the minimum, which would add additional time to raise the needed equity.

[6] The cost of capital is a sophisticated financial concept. In colloquial terms, it is the rate of return required by investors which results in a market value of a company’s equity being “at book”, meaning what was paid for it (adjusted for any retained earnings since). One can closely estimate it for a publicly-traded stock. As the GSEs’ stocks, while they exist, do not trade based upon economic fundamentals, it is necessary to look at what other roughly comparable financial institutions with well-traded shares have as their calculated cost of capital. In conservatorship, the FHFA has a list of such comparable organizations and uses it to set G-fees under their authority as the conservator of the two companies.

[7] This is referring to the average G-fee. There is a system of (1) risk adjustment of the fees, and of (2) cross-subsidies (set by the FHFA) so any individual loan might have a G-fee considerably higher or lower than the average. Congress has also added a 0.10% additional tax to the G-fee that goes straight to Treasury. Currently, the total average G-fee is about 0.58% (which includes the tax).

[8] There were lots of different groups, with different motivations, who wanted to see the exit happen much sooner. This included the GSE managements and Boards wanting to be freed of government control; the Wall St. firms hoping to earn large fees from underwriting new share issuances; and government officials and policymakers who wanted the “last unresolved issue from the Financial Crisis” (as it is sometimes called) to finally be resolved and put to bed. Among the last group was FHFA Director Calabria who, early in his tenure, talked about doing GSE equity issuances within a year or so after his arrival in 2019, while the companies would still be in conservatorship, to hasten its end; needless to say, nothing came of it.

[9] It is theoretically possible Treasury could just waive its right to exercise the warrants. But I do not believe it is a real possibility - to do so would be an immense give-away of taxpayer money, and thus politically highly controversial as a result.

[10] The current outstanding amount is $222 Billion.

[11] That precedent is the government rescue of AIG, where Treasury ended up, immediately post conversion, owning over 90% of the shares.

[12] If Treasury tries to sell too quickly, the result is that the shares of the GSEs would become very depressed, which could flow into making the operations of the two companies significantly more difficult.

[13] It is not as simple as Treasury owning less than 50% of the shares of each GSE, as it would still be by far the largest shareholder at such a point. So, it is somewhat a judgment call as to what percentage would result in Treasury control, in practical terms, ceasing. Only at very small percentages, or down to fully zero, would it be certain Treasury was no longer in control. However, Treasury would still be providing financial support to the two companies via the PSPA agreements, which would give it certain rights over the two companies, although nothing like operating control.

[14] If conservatorship, and its companion PSPA, last past 20 years, the PSPA will likely need to be amended in certain ways to accommodate it evidently lasting longer than the drafters of the PSPA ever contemplated.

[15] See my previous article showing the low level of losses predicted by the stress tests, from August 2021 “The GSE Stress Test Results: Good News but Troubling Decisions.” https://www.jchs.harvard.edu/blog/gse-stress-test-results-good-news-troubling-decisions.

[16] Less well understood is that a regulatory capital rule has to also be well-constructed in terms of supporting economically sensible risk-versus-reward decisions. That is revealed not in the grand total of capital required, but in the details of how much required capital grows or shrinks in response to specific transactions. The current capital rule is considered flawed in that respect, with the previously-discussed change in how CRT is handled an example of trying to improve the rule. Getting this right is another reason why a new capital rule is required.

[17] From 2008 through late 2012, via a 10% coupon on the amounts utilized; from 2012 to 2019, via large dividend payments that swept almost all the earnings of the GSEs to Treasury; and since 2019, via Treasury receiving quarterly more shares of preferred stock in the two companies with a face amount basically equal to their earnings for the quarter.

[18] The fee can be a flat fee or one that varies with the riskiness of the GSE portfolios. It can also be structurally based upon outstanding guarantees or upon total assets or even upon the unused amount of funding available under the PSPA.

[19] All such regulation uses one version or another of promising that the rate-setting will allow investors in the utilities (and other regulated industries) to earn a “fair” return. There can naturally be disagreement between regulator and regulated about how to define and calculate “fair.”

[20] For example, (1) excessive travel & entertainment expenses, or (2) lobbying expenses.

[21] As a practical implementation tip, the FHFA can hire experts in state utility regulation to do this.

[22] Government agencies are subject to various restrictions that include, as examples: (1) having their budgets set by Congress, (2) being subject to government compensation schedules and practices, and (3) having specific directions, requirements and limitations given to them by Congress and/or the White House that often reflect political considerations. All these mitigate against a typical private-sector level of competitiveness and innovation.

Donald H. Layton

Donald H. Layton is a Senior Visiting Fellow from Practice. Prior to joining the NYU Furman Center, he served as a Senior Industry Fellow at Harvard’s Joint Center for Housing Studies, where he wrote extensively about the Government Sponsored Enterprises (GSE) of Freddie Mac and Fannie Mae and more broadly on housing finance. Before his stint in academia, Layton was the CEO of Freddie Mac from May 2012 until June 2019, where he championed the development of Credit Risk Transfers, one of the most significant reforms to the housing finance system in decades.

« Previous | The Stoop | Next »