The Heavy Lift to Implement GSE Reform-Recap-Release: Recapitalization (Part 3 of 4)

One of the most discussed topics in housing finance policy in recent months has been whether and when the second Trump administration (Trump II) might end the 16-year conservatorships of Fannie Mae and Freddie Mac (F&F), the two large government-sponsored enterprises (GSEs).  This four-part series examines what it will take to actually implement that exit, a process known informally as “reform-recap-release.” [1]

Part 1 of this four-part series gave the necessary background, introducing two fundamentally different types of reform to address. Part 2 discussed the first of those two types: reforms needed to correct what many policymakers have long deemed to be significant defects in the charters of F&F, several of which were proximate causes of their falling into conservatorship. 

Now, in Part 3, I address recapitalization through two major workstreams: (1) develop and implement a plan for F&F to have enough capital to qualify conservatorship exit, and (2) develop and implement a plan for Treasury to dispose of its large ownership interest in F&F.  While discussed as two separate workstreams, both must in fact be integrated into a coherent whole, or potentially one or both will fail. [2]

Workstream 5 [3] – Primary capital raise: Develop and implement a plan for F&F to have enough capital to qualify for exit from conservatorship.

This workstream addresses three related questions: (1) What amount of capital is needed to qualify for conservatorship exit? (2) How long will it take for earnings retention alone to reach that level of capital? and (3) Is the issuance of new shares a good option for F&F to recapitalize sooner than if done solely by retaining earnings?

            What amount of capital is needed to qualify for conservatorship exit? F&F each need to have a level of capital [4] that meets or exceeds the requirements set by their regulator, the FHFA, to ensure safe and sound operations. [5] In addition, it is important for them to have a buffer above the regulatory-determined minimum. Such a buffer will help protect against losses from a possible economic downturn causing a breach of that minimum, which would in turn trigger unwanted regulatory sanctions. Based on research done with the FHFA circa 2018, a cushion of an additional five percent seems reasonable. [6] As already discussed in Part 2, the current regulatory minimum, based upon formulae referred to as the Enterprise Risk Capital Framework (ERCF), generates a requirement of about $350 billion, inclusive of the five percent cushion, and acts as a “high” estimate. The “middle” estimate, which builds on the previous Conservatorship Capital Framework while also incorporating a countercyclical buffer (the enhanced CCF), generates a requirement of about $230 billion, also inclusive of the buffer. As already recommended in Part 2, the FHFA should take a fresh look at possibly replacing the ERCF, as many (myself included) believe it is set too high. This perception stems from how inconsistent it is with the results of the government-designed stress tests on F&F, which show little to no loss.

            How long will it take for earnings retention alone to reach that level of capital?  In the last three years, the average net income for F&F was $26.3 billion. However, GSE earnings are sensitive to economic conditions, in particular changes in home prices, all of which have been very favorable in recent years. Thus, for planning purposes, a slightly more conservative estimate of net earnings of $20 billion per year seems reasonable going forward. As of the end of 2024, the combined net worth of F&F was $154 billion. That means it will take roughly another 10 years to reach the high estimated capital requirement of $350 billion, but only about four years to reach the middle estimate of $230 billion. [7] In the case of the 10-year time horizon, a presidential administration clearly has to plan on implementation extending to future presidencies, along with all the uncertainties that come with such a long period of time. In contrast, a four-year time horizon presents fewer uncertainties and can also be largely implemented during a single presidency. This is a significant difference and indicates why it is so important for the FHFA to review its capital requirement to determine if the current ERCF really is appropriate or is instead too high.   

            Is the issuance of new shares a good option for F&F to recapitalize sooner than if done solely by retaining earnings? Issuing new shares to help accelerate when F&F’s capital will be high enough to enable conservatorship exit has been discussed for many years. And while the concept of selling new shares seems sound at a general discussion level, it actually is highly problematic for reasons related to the particular nature of conservatorship and the unusually large size of F&F. [8]

• First, issuing new shares can only accelerate the timing of full recapitalization by roughly one year, i.e., from 10 to nine years for the high capital estimate and from four to three years for the middle estimate. This disappointing result stems from a combination of several unusual factors:
  • Raising equity during conservatorship is basically a non-starter, as potential investors would reject becoming shareholders of F&F knowing that they would be totally disenfranchised, i.e., with no ability to receive dividends, vote for board members, etc. However, if F&F are to exit conservatorship safely and soundly, they need to be fully capitalized upon exit. That means the only stock issuance that is both possible and relevant is one that closes both simultaneously with, and also contingent upon, an exit from conservatorship.
  • This timing then presents an additional obstacle: practical limits on how much could be raised in a single equity-raising transaction that closes upon conservatorship exit. Historically, two “largest” stock issuances stand out: the largest global issue was $29.4 billion for Saudi Aramco, while the largest U.S. company stock sale was the initial public offering of VISA at $17.9 billion. In fact, stock issues above $10 billion are fairly rare. Thus, F&F can be expected to raise at most roughly $20 billion in a one-and-done capital raise that would close upon conservatorship exit, far less than what is needed to fill the entire shortfall – about $200 billion in the high capital estimate, and $75 billion in the middle one.  In fact, given the annual earnings planning assumption made above of also about $20 billion per year, such an equity issue will shorten by only about one year the estimated 10-year and four-year timeframes for fully building capital from retained earnings. That’s not much of a difference, and pursuing this option could have very negative side effects, as I will discuss below.
• Second, the stock price will likely have to be highly discounted to attract equity investors, meaning that the new stockholders would own a disproportionately large share of the company compared to the dollars they invested. [9] That’s because the uncertainties surrounding the revenues and profits of F&F after conservatorship versus what they have been during conservatorship are unusually high, especially given how much the government and politics impact the two GSEs. [10] Such unusually high uncertainties will likely be front of mind for investors in a potential equity share sale, pushing the price of the shares down to where they are very strongly discounted. This could create two key undesirable side effects.
  • It will force the existing shareholders – who are mainly U.S. taxpayers through Treasury – to take a giant loss as their shares get very heavily and permanently diluted. 
  • A discounted share price also raises the cost of capital, which in the case of F&F will feed right into pressure to raise G-fees, as the cost of capital is perhaps the largest cost that goes into setting G-fees. 

Added up, it therefore seems unwise to pursue a share issue as part of a plan to recapitalize the GSEs to the level needed for conservatorship exit:  the upside of shortening the time of conservatorship exit by about one year is very limited, and the downside is likely quite damaging. [11] 

Workstream 6 – Treasury’s ownership disposition:  Develop and implement a plan for Treasury to dispose of its large ownership interest in F&F. 

F&F were rescued under terms whereby (1) Treasury received a 79.9 percent warrant on their common equity shares, [12] thereby diluting existing shareholders’ ownership interest in the company by nearly 80 percent, plus (2) any funds used to inject equity into F&F to maintain mortgage market confidence in them – an unknown amount when conservatorship began in 2008, but which grew to $191 billion over the next three-plus years – were in the form of a senior preferred shares investment. This combination has left the common shareholders with a heavily diluted investment, which is also, of course, junior to the preferred shares held by Treasury. [13]

It is beyond the scope of this paper to go through the many evolutions of the PSPA agreement since 2008, but some history is absolutely necessary to understand several issues related to Treasury eventually disposing of its ownership interest. Starting in 2012, the original 10 percent dividend on the senior preferred stock was replaced by a “sweep” dividend of almost all F&F’s earnings – a highly unusual move that not only prevented the GSEs from recapitalizing but actually forced their decapitalization. This unsurprisingly led to various lawsuits alleging that this constituted an unfair seizure of value from the public shareholders who had already been disenfranchised during conservatorship. While the government won almost all the lawsuits, the bad feelings on the part of those shareholders – by that time, mostly professional investors making a highly speculative investment – did not go away. In 2019, during Trump I, those shareholders argued that the senior preferred stock should be simply cancelled in an exit, since the sweep dividends had, it turned out, proven to be much higher than the $191 billion invested by Treasury. Next, when the so-called sweep dividend was ended in late 2019, F&F began to retain their earnings to build capital.  However, in another unusual twist, instead of F&F paying a preferred shares dividend to Treasury, the senior preferred stock increases dollar-for-dollar by the amount of the earnings retained; this was designed to ensure that the earnings retained belonged wholly to the taxpayer and not partially to any private sector investors in the common.

Today, taking into account the twists and turns in the PSPA described above, Treasury holds warrants on 79.9 percent of the common stock of each of F&F and also owns senior preferred stock with a face value (technically known as the “preference”) of $341.0 billion. 

With this background, this workstream is designed to answer two questions: (1) Should Treasury cancel the outstanding senior preferred stock or alternatively convert it to common stock? and (2) What would be a reasonable plan for Treasury to dispose of the common shares obtained through exercising the 79.9 percent warrants and any shares from a conversion of the senior preferred stock? 

            Should Treasury cancel the outstanding senior preferred stock or alternatively convert it to common stock?  This is overwhelmingly an intermixed policy and political decision. On the policy front, such an action would benefit the shareholders, who would have otherwise lost everything if F&F had failed back in 2008. When this issue was discussed during Trump I, Treasury Secretary Steven Mnuchin spoke firmly against just such a cancellation. On the political front at the time, almost all Democrats in Congress, and many Republicans as well, immediately labeled this possible action as an unacceptable giveaway of the taxpayers’ money.  The proposal went nowhere. 

Nevertheless, the issue is back on the table because those same Wall Street professional investors are advocating again for such a cancellation. Trump II officials have not publicly commented on the issue to date. There is also no obvious reason to believe that Congress’ past strong opposition to such a cancellation has changed in the intervening years. 

However, if not cancelled, what to do about the $341 billion outstanding? The only precedent is the AIG bailout and its resolution, [14] which occurred between 2011 and 2013. At that time, Treasury converted the senior preferred stock into common stock, allowing the resulting common shares to be sold over time into established and relatively liquid markets for AIG common stock. In my view, this approach seems to be the only practical alternative for F&F as well. [15] 

            What could be a reasonable plan to dispose of the common shares obtained through exercising the 79.9 percent warrants and any shares from a conversion of the senior preferred stock?  The decision to sell such a large number of shares – a very significant amount both in terms of dollar value (likely multiples of the largest common stock sale ever) and ownership stakes in each of F&F – should be made by Treasury based on technical market considerations after presumably being advised by equity market experts. It could perhaps take the form of a series of large, underwritten “secondary” sales from time to time, or perhaps selling modest amounts every day, or anything in between. Regardless, to avoid pushing strongly down on the price of the common shares, sales will need to be spread out over a large number of years.

However, before undertaking any such sales, a major policy decision is required. The very definition of conservatorship is that the government, currently through the FHFA, is in operational control of F&F, rather than their shareholders. While this will end when the GSEs exit conservatorship, the reality is that the government will still be in operational control of F&F, but instead through Treasury as the controlling majority owner of the companies’ shares.  Some dub this a form of “conservatorship by other means.”  The publicly owned common shareholders will thus still be significantly disenfranchised, and the shares will undoubtedly trade at a lower price to reflect this.  This means the share sales by Treasury will produce far less revenue than might be expected. 

In the case of AIG, which as stated above is the only obvious and relevant precedent, Treasury voluntarily gave up almost all of its majority shareholder control to avoid a discount on the price of the common stock and speed up sales. It did this by agreeing to vote its shares exactly pro rata with the votes of publicly-owned shares, except for a few core governance items. For F&F, doing something similar would mean that government control of the two companies would almost totally end upon conservatorship exit, and that public shareholders would no longer be a disenfranchised minority. The impact is that the share price would clearly be higher, producing greater revenues for Treasury over time as it sells its shares. But the administration will have to accept giving up that operational control of F&F, which, after nearly two decades, might be politically harder than it should. [16]

Conclusion

As described in this piece, the administration will have to make significant and controversial decisions as it develops a comprehensive plan to (1) recapitalize F&F enough to qualify for conservatorship exit, and then (2) for Treasury to best dispose of its very large – both in dollar and percentage terms – ownership interest. For example:

• Lowering the regulatory minimum capital requirement from today’s level would be controversial – and will likely be opposed by small government conservatives. 
• Deciding to use retained earnings alone to reach full recapitalization would likely be opposed by the investment banks and others who hope to earn large fees from a stock sale.
• Converting the senior preferred stock to common stock, rather than cancelling it, would also be very strongly opposed, possibly also via lawsuits, by the professional investors in the common shares.
• The administration giving up the operational control of F&F that would otherwise accompany Treasury’s majority ownership post-conservatorship, as was done with AIG, would also be controversial, and probably opposed by groups that like greater political control of the two GSEs.

Lastly, undertaking a complete recapitalization, including disposing of Treasury’s ownership interest, would represent a rather large commitment of time and political capital from Trump II’s senior policymakers. Alternatively, rather than a full-bore effort at conservatorship exit, Trump II could instead focus on making targeted progress in key areas, which will in turn make it easier for a future presidential administration to finish the job.  That is, after all, exactly what happened during Trump I. 

Part 4, the last instalment of this series of papers, will address the second type of reforms identified in Part 1: those driven by the FHFA, using its conservatorship authority over F&F, to make the mortgage markets work better. Such reforms, never expected at the beginning of the conservatorships and still not well understood as a comprehensive program, have nevertheless become embedded widely throughout the mortgage markets – but are at risk of messily being unwound if not thoughtfully addressed ahead of time as part of conservatorship exit planning. 

Footnotes

[1]As already described in Parts 1 and 2, this series is based upon the commonly-held view that Congressional legislation about GSE reform will not occur in any reasonable timeframe, and that its implementation will be accomplished solely through administrative means, meaning that there will be no legislation involved. Instead, all the changes needed for implementation will be made via either a Federal Housing Finance Administration (FHFA) regulation, but only for those topics related to safety and soundness, or a clause in the Preferred Stock Purchase Agreement (PSPA), the legal agreement by which Treasury provides credit support to F&F. 

[2] I bring to this discussion of recapitalization more than just my expertise as the former CEO of Freddie Mac from 2012 to 2019. In my previous banking and finance career, I had critical experiences that are relevant.  First, I was the head of trading and sales activities for one of JPMorgan Chase’s predecessors starting in 1988, and continued in that position through the mergers that created JPMC, including being co-head of its investment bank, through 2002. These activities included public equities in their later years, in addition to the company’s long-standing large fixed income, foreign exchange, and other businesses. On a transaction basis, I had personal involvement with two major recapitalization equity issuances, one related to the merger of Manufacturers Hanover Corporation with the Chemical Banking Corporation (where I led the international “roadshow”, i.e., meetings with potential investors to develop interest in purchasing new shares), and the other by E*TRADE Financial (where I was Chairman & CEO).  Also of direct relevance, Treasury placed me on the board of AIG in 2010, where I was able to participate in its recapitalization after the 2008 government rescue, which followed the same financial format as F&F (i.e., 79.9 percent warrant plus senior preferred shares) and is the only extant precedent for a recapitalization of this type and size being completed. 

[3] The first four workstreams were in Part 2 of this series. These are numbered sequentially through the four parts.

[4] In this series, net worth will be used as a practical proxy for capital.

[5] All regulators of financial institutions have formulae to determine how much capital a particular company needs to operate on a safe and sound basis, at a minimum, given the risks it takes. 

[6] There is a view that F&F could possibly be released from conservatorship prior to them meeting their full regulatory capital requirement via a regulatory mechanism called a “consent decree.” This view is unconventional and not broadly held; thus, this alternative has not been included in this series of papers. 

[7] The estimates above of four and 10 years are just a simple extrapolation. It may be worthwhile to develop a more sophisticated model to estimate the number of years in each case that includes assumptions about growth in the volume of mortgages originated. 

[8] F&F are, in some sense, alter egos for the U.S. government. As such, they are extremely large when compared to private sector financial institutions. For example, the largest U.S. bank is JPMorgan Chase, which had yearend 2024 assets of $4.0 trillion, whereas F&F together had $7.5 trillion. 

[9] For example, if F&F had $225 billion of capital but needed $250 billion for a conservatorship exit, they could raise an additional $25 billion, which is 10 percent of the new total capital amount. But if the shares were issued at a discount, the new shareholders would own not a pro rata 10 percent of the company’s shares going forward but perhaps 20 percent, 40 percent, or even possibly more. 

[10] Here are just three examples of such uncertainties: (1) G-fees are now set by the FHFA; it is unclear how they will be set post-exit or whether that would lead to them being higher or lower; (2) F&F may have to pay a fee for Treasury support to the companies, and it is unknown how it will be set, what level it will be, and also how easily the government could change it in the future; and (3) Congress currently charges hidden taxes on F&F to fund various housing programs that, together with the two companies paying for FHFA expenses, amounted to roughly half their total operating expenses in 2024 – and Congress could easily add more such levies on the companies in the same tradition.

[11] Nevertheless, the Trump administration will face an onslaught of investment bankers and others recommending and lobbying to do just such an issue in order to earn the very large fees associated with doing so.

[12] The major reason the government did not take a warrant on more than 79.9 percent of F&F’s common shares was to prevent F&F’s debt – which at the time amounted to over $4 trillion – from being consolidated into the federal budget, which would have meant a $4 trillion increase in the federal debt and subsequent debt ceiling problems. 

[13] F&F also have outstanding preferred shares issued in the public markets prior to conservatorship, now often referred to as the “junior preferred,” with $33.2 billion outstanding. In conservatorship, their dividends were and remain suspended, and their position is, of course, junior to that of Treasury. 

[14] Treasury placed me on the board of AIG in 2010 as part of its rescue. I served in that role until becoming CEO of Freddie Mac in May 2012. For AIG’s bailout, Treasury used the same 79.9 percent warrant/senior preferred stock structure as it did for F&F.

[15] There is one additional alternative: for Treasury to be paid back the senior preferred stock over time, earning a dividend meanwhile. However, this is basically a financial non-starter because the amount of the senior preference at $341.0 billion is so large. First, the dividend would drain the companies and almost wipe out their ability to build capital or pay dividends (e.g., a five percent senior preferred dividend rate, or about $17 billion per year, would take almost all of the earnings of the two companies). Second, the amortization schedule to pay back the principal would go on for many decades as F&F simply can’t afford anything faster – unless, of course, G-fees were significantly increased, which is itself objectionable to most policymakers, in my experience. Choosing this type of path, i.e., where the senior preferred stock is slowly paid back and takes most earnings for its dividends, would also strongly hinder efforts to raise common equity.

[16] As described in Part 2, the government can include clauses in the PSPA that keep various types of approval and decision-making authority for itself even after conservatorship has ended. If that is done too extensively, however, it will create exactly the kind of “conservatorship by other means” discount in the share price that the loss of Treasury’s majority control is designed to avoid.