Shortly after my book The Mortgage Wars came out in November of 2013, an executive of the Fairholme Funds connected me with the legal team at Cooper & Kirk in Washington D.C. (about 15 miles from my home), whom Fairholme had retained to pursue its suits against the government for imposing the Fannie and Freddie net worth sweep in August of 2012, and subsequently I flew to Miami for two days of meetings at the Fairholme offices. Thus began my involvement with all of the net worth sweep cases, the first filed on July 7, 2013.
I had written about the sweep in my book. I had noted that “In the first year and a half after its conservatorship, Fannie Mae reported a staggering $127 billion in losses, exhausting its capital and causing it to draw $75 billion under its senior preferred stock agreement with Treasury in order to maintain a positive net worth.” I pointed out, however, that of this $127 billion only $16 billion were actual credit losses, and that, “Virtually all of the rest of its losses were accounting entries made by the company’s conservator, FHFA, that pulled into Fannie Mae’s 2008 and 2009 financial statements over $100 billion in expenses that otherwise would have been incurred in the future, if they were incurred at all.” Fannie’s draws of senior preferred rose to $116 billion at the end of 2011, but then, after home prices bottomed out, it began to make money again. Fannie announced on August 2, 2012 that it had earned $5.1 billion in the second quarter, enough to add $2.5 billion to its net worth after the required senior preferred stock dividend to Treasury. Less than ten days later, Treasury and FHFA agreed to the net worth sweep, changing the dividend on Fannie’s senior preferred stock from 10 percent per annum to “everything you ever earn.”
I understood exactly what had happened. Treasury knew that a large number of temporary or estimated non-cash expenses put on Fannie’s books by FHFA were about to reverse and come back into income, not only greatly increasing the company’s capital but also making clear that its 2008-2011 losses had not been real, and that Fannie had been forced to take senior preferred stock it didn’t need but couldn’t repay, whose purpose was to “transform massive, temporary and artificial book expenses created for Fannie into massive, perpetual and real cash revenues for Treasury.” I laid out my thesis in a January 2015 paper titled Treasury, the Conservatorships and Mortgage Reform (accessible on this site, under the column labeled “Reference Documents”), and six months later, at the request of Cooper & Kirk, expanded on this paper in an amicus curiae brief written for the Perry Capital case, being heard by Judge Royce Lamberth. I did a second amicus brief in February of 2016 for the Jacobs-Hindes case in Delaware, and discussed it in this blog’s inaugural post, titled “Thoughts on Delaware Amicus Curiae Brief.” Back then I believed it was inevitable that the net worth sweep would be invalidated in the courts, as I said in that post’s final paragraph: “Sorry, Treasury. Because of the lawsuits, you’re now in a different game. Your actions, and your defense of those actions, no longer are being adjudicated only on the editorial pages of the Wall Street Journal; they also are being adjudicated in courts of law. There facts matter, and there you almost certainly will lose.”
Seven years later, those words ring hollow. Through some two dozen lawsuits contesting the net worth sweep, counsel for Treasury and FHFA have continued to advance a blatantly and provably false version of the reasons for and the actions behind it, with no significant adverse legal consequences. On the one case that reached the Supreme Court, Collins v. Yellen, the author of its unanimous opinion, Justice Alito, repeated the government’s false version of the rationale for the sweep—ignoring the contrary factual allegations put forth by the complainant (which in a motion to dismiss must be accepted as true), as well as the amicus brief I filed with the Court—then insisted that a clause appearing in a section of the Housing and Economic Recovery Act (HERA) titled “Incidental Powers,” allowing FHFA to act “in the best interests of…the Agency,” could be read to apply to the statute as a whole, thus overriding the specific “Powers of Conservator” listed in an earlier section. And even here, Alito had to construe that giving all of Fannie and Freddie’s future earnings to Treasury somehow was “in the best interests of the Agency [FHFA]” in order to rule that the net worth sweep was a legal act by the conservator.
For me, the Alito ruling was a defining event. I of course knew that he and the other five conservative justices on the Roberts Court were members of the Federalist Society, which was founded in 1982 and has been a fierce opponent of Fannie and Freddie since that time. But still, I did not believe the Court would make such a nakedly political ruling in the net worth sweep case until it actually did. Yet now that it has, it is foolish to ignore the ruling’s implications. One is that none of the remaining constitutional challenges to the net worth sweep percolating in the lower or appellate courts—whether on the succession clause, the appointments clause or the separation of powers—has any realistic chance of resulting in a voiding of the net worth sweep that is upheld by the Roberts Court. While it is possible that plaintiffs in suits alleging regulatory takings or breach of contract may win some minor amount in monetary damages—and even that will not be easy—the net worth sweep, as flagrantly illegal as it was, will not be overturned judicially.
With the net worth sweep legitimized, and a capital standard that imposes bank-like requirements on Fannie and Freddie’s low-risk credit guaranty business by adding huge amounts of conservatism, minimums and buffers to the risk-based framework mandated by HERA—and pretending that the companies’ $35 billion per year in guaranty fees absorb no credit losses—the companies are in an impossible position. The $192 billion in senior preferred stock that they were forced to take, could not repay, and doesn’t count as core capital must remain on their balance sheets, and they must replace the $265 billion in earnings swept by Treasury between the end of 2011 and the middle of 2019 with earnings they’ve been allowed to retain since then, at the cost of an equal increase in Treasury’s liquidation preference, now $289 billion and rising every quarter. And before they can be considered adequately capitalized under the standard made final by former FHFA Director Calabria in December of 2020, they must attain a level of “adjusted total capital” that as of September 30, 2022 was $301 billion, or 4.0 percent of total assets, a percentage that likely will increase in the future because it is procyclical.
This untenable construct for the net worth sweep and the “Calabria capital standard” was hard-wired into the January 14, 2021 letter agreement between former Treasury Secretary Mnuchin and Calabria. That letter stipulates that neither Fannie nor Freddie will be eligible to be released from conservatorship until “all material litigation [is] resolved or settled,” and each company “for two or more consecutive calendar quarters has and maintains ‘common equity tier 1 [CET1] capital’…in an amount not less than 3 percent of Seller’s [Fannie’s or Freddie’s] ‘adjusted total assets’.” Subsequently, the companies may continue to retain capital until “the last day of the second consecutive fiscal quarter during which Seller has had and maintained capital equal to or in excess of all of the capital requirements and buffers under the Enterprise Regulatory Capital Framework [ERCF]”. At that point, the net worth sweep will be resumed, at an amount “equal to the lesser of 10.0 percent per annum on the then-current Liquidation Preference and a quarterly amount equal to the increase in the Net Worth Amount, if any, during the immediately prior fiscal quarter.”
The letter agreement does permit each company to issue up to $70 billion in common stock (the same amount for both, notwithstanding that Fannie’s required capital is half again the size of Freddie’s) provided all material litigation has been resolved or settled, and Treasury has exercised its full warrant to acquire 79.9 percent of that company’s common stock. But few if any investors will put new capital into the companies, much less $70 billion, while the net worth sweep remains in place and Treasury’s liquidation preference rises in line with their earnings. Consequently, as long as the January 14, 2021 letter is in effect, Fannie and Freddie seem destined to remain in conservatorship until they can accumulate CET1 capital equal to 3.0 percent of their adjusted total assets, and they will have constraints on their executive compensation and dividend-paying abilities until they fully meet “all of the capital requirements and buffers” of Calabria’s ERCF.
How long might that take? We can make an estimate, using data as of September 30, 2022 from the companies’ most recent 10Qs. To be eligible to be released from conservatorship, Fannie would need $136 billion in CET1 capital, $230 billion more than the negative $94 billion it has now (because of the net worth sweep); Freddie would need $111 billion, or $168 billion more than the negative $57 billion it has. At what I estimate as their sustainable rate of after-tax retained earnings—about $13 billion per year for Fannie and about $9 billion per year for Freddie—Fannie would not be eligible for release until 2040, and Freddie not until 2041. Fully meeting their risk-based capital requirements would take even longer.
After over 14 years in conservatorship, what possibly could justify keeping the companies in conservatorship for another 18 years, or more? There is no economic reason. Fannie and Freddie are, and always have been, the highest-quality and lowest-risk sources of mortgage credit in America. Prior to the 2008 financial crisis, the serious delinquency rate on single-family loans owned or guaranteed by the companies was one-third that of the prime single-family mortgages held by banks, and less than one-tenth that of the mortgages originated by subprime lenders. Along with the rest of the industry, Fannie did experience a spike in credit losses on loans guaranteed between 2004—the year the issuance of private-label securities first exceeded the combined issuance of mortgage-backed securities (MBS) by Fannie and Freddie, and underwriting standards collapsed—and 2008, suffering average lifetime default rates on these books of about 10 percent (comparable data for Freddie are not publicly available). But from 2009 on, each of Fannie’s books of business has a lifetime default rate of less than 1 percent. And Fannie and Freddie’s total single-family credit guaranty books as of September 30, 2022 are pristine, with average current loan-to-value ratios of 50 and 53 percent, and credit scores at origination of 752 and 750, respectively. Any doubts about the companies’ credit quality should have been dispelled by the results of their last two Dodd-Frank stress tests, in which neither needed a dollar of initial capital to withstand a stylized repeat of the 30 percent nationwide decline in home prices during the time of the Great Financial Crisis.
Nor is there any structural reason to keep Fannie and Freddie under government control. The most severe and persistent criticism leveled against the companies prior to the crisis was that they should not be allowed to finance mortgages on their balance sheets; today they no longer are. And their entity-based credit guaranty business does not need “reform,” because it already is far superior to the senior-subordinated model used in private-label securitization. In the entity-based model, revenues on good loans from all years, regions and loan types are available to cover losses on any loans that go bad, and all loans benefit from a corporate guaranty. In contrast, each senior-subordinated pool must stand on its own, and the inability to reach beyond it for revenues—or add capital post-securitization—requires substantial initial subordination, which translates into considerably higher credit guaranty costs, while still leaving investors in the senior tranches exposed to losses that exceed the fixed loss-absorbing capacity of the subordinated tranches. Fannie and Freddie’s entity-based model makes them the credit guarantor “gold standard,” as they exist today.
Fannie and Freddie have neither an economic problem nor a structural problem. They have a political problem. And it will require a political solution.
It was a political judgment that put Fannie and Freddie into conservatorship in the first place. After the private-label securities market imploded in 2007, and banks began pulling back sharply on their mortgage lending, Treasury Secretary Hank Paulson knew this left Fannie and Freddie as “the only game in town” (his words) for mortgage lending. Because he did not want to rely on them as shareholder-owned entities to get the country through the financial crisis, however, he made the policy decision to nationalize them, while each exceeded their statutory capital requirements. Except he couldn’t admit that; it had to be a “rescue.” Thus was born the litany of fictions about Fannie and Freddie that persist to this day: that the $187 billion in senior preferred they had been forced to take (to cover non-cash losses) was a real economic cost, and a “bailout;” that they—and not the private-label securities market—had caused the financial crisis, and that their “flawed business model” required them to be “wound down and replaced” by Congress in legislation.
Congress, as we know, has been unable to come up with a workable alternative to Fannie and Freddie, because there isn’t one. What I call the Financial Establishment—large banks and Wall Street firms, and their advocates and alumni at Treasury and elsewhere—finally has accepted this, but continues to insist that Fannie and Freddie only can be released from conservatorship if they are capitalized like banks, and “reformed” (with the two preferred proposals being an explicit government guaranty on their securities, and allowing FHFA to charter multiple credit guarantors). Former FHFA Director Calabria obligingly saddled Fannie and Freddie with a meticulously engineered “risk-based” capital standard that produces 4-percent-plus capital requirements for the companies no matter how little risk they take, while current FHFA Director Thompson says she will defer to Congress on whether, how and when to move Fannie and Freddie off their 18-year-plus path of exiting conservatorship through retained earnings alone, with the net worth sweep and the Calabria capital standard staying as they are in the meantime.
It is hard to imagine, though, that Fannie and Freddie will remain in counter-factual limbo until 2040. Well before then, it seems inevitable that some administration will recognize and admit the absurdity of the status quo, and acknowledge the damage it is doing to the nation’s housing finance system. Whichever administration does so will get to restructure, recapitalize and release the companies in the manner and with the objectives it chooses.
The Biden administration is first in line. And Fannie has not been shy about expressing the extent of its problems with the policy and regulatory actions of FHFA and Treasury. In its 2021 10K, for example, Fannie said, ”We believe that, if we were fully capitalized under the [ERCF] framework, our returns on our current business would not be sufficient to attract private investors,” and that, “Increasing our returns may require substantial increases in our pricing or changes in other aspects of our business that could significantly affect…the level of support we provide to low- and moderate-income borrowers and renters.”
The average guaranty fee rate on Fannie’s single-family book of business nearly doubled between 2007 and 2021, rising from 23.7 basis points to 45.2 basis points (not including the 10 basis points it has had to charge and remit to Treasury since April of 2012). During this same period, Fannie’s percentage of credit guarantees on loans to borrowers with credit scores under 700 fell from 33 percent at December 31, 2007 to less than half that, 15 percent, at December 31, 2021. Fannie’s guaranty fees in 2021 already seemed to be at the threshold of affordability for most lower-credit-score borrowers, but the company’s need to get closer to a market rate of return on its required ERCF capital forced it to raise its average fee on new business in the third quarter of 2022 to 53.3 basis points (63.3 with the 10 basis points for Treasury), with that fee likely headed even higher, and the percentage of Fannie’s business done with lower-credit-score borrowers headed lower.
It would be a simple matter for a Democratic administration with a stated policy priority of supporting affordable housing to remedy this situation. It merely has to declare that the 2008 nationalization of Fannie and Freddie (by a previous administration) was unjustified and a mistake, and that it has had the severe and ongoing consequence of impeding access to homeownership for low- and moderate-income families. To reverse that action, it would declare the companies’ senior preferred stock to have been repaid (which it has been) and cancel it, along with Treasury’s liquidation preference. Next, it would require that Director Thompson (or her successor) replace the Calabria standard with one that is not designed to produce a pre-determined capital number, but instead reflects the actual risk of the loans the companies are guaranteeing, with a minimum required capital percentage consistent with those risks. In Capital Fact and Fiction, I explain why that minimum should be 2.5 percent, and note that with the credit quality of Fannie and Freddie’s current business, the 2.5 percent minimum would be their binding capital percentage for the foreseeable future. Eliminating Treasury’s senior preferred stock and liquidation preference, and a putting in place a true risk-based capital standard with a 2.5 percent minimum, will create a path for Fannie and Freddie to emerge from conservatorship relatively quickly, and to resume their traditional roles as the mainstays of large-scale, low-cost, finance for affordable housing.
Standing in the way of this remedy, however, are the Financial Establishment and its ally in the judiciary, the Federalist Society, who continue to falsely claim that the conservatorships of the companies are justified, the net worth sweep is proper compensation for Treasury’s “heroic rescue” of them, and that they only can be released if they are capitalized like banks and “reformed.” Some of the reason for this posture on Fannie and Freddie is ideological, but most is competitive. Banks have benefited greatly from having Fannie and Freddie run in conservatorship and grossly overcapitalized by FHFA. At December 31, 2007, banks held $2.29 trillion in single-family whole loans and MBS, or 23 percent of outstanding single-family mortgage debt, on their balance sheets. At June 30, 2022 (the latest date for which full data are available), banks held more than double that amount—$4.65 trillion, for a 36 percent market share. Moreover, because the rates on all new mortgages are set with reference to MBS yields, the same unnecessarily high guaranty fees that are blocking access to mortgage credit for affordable housing borrowers add, basis point for basis point, to the spreads on the long-term fixed-rate whole mortgages banks finance in their portfolios with government-guaranteed consumer deposits and purchased funds. It is, as they say, “about the money.”
Yet this cliché also highlights a compelling argument for the Biden administration to not keep giving the banks what they want, and instead “do the right thing” by Fannie and Freddie, and low- and moderate-income homebuyers. After Treasury told FHFA to put Fannie and Freddie into conservatorship, it gave itself warrants for 79.9 percent of each company’s common stock. Today, those warrants have very little value, just $3.2 billion, because—with the government’s current policy of laying claim to more than all of Fannie and Freddie’s net worth, and keeping them unreasonably mired in conservatorship for the next two decades—the companies have very little value, with their share prices averaging 44 cents yesterday. As noted earlier, I estimate their combined sustainable earnings to be about $22 billion per year. At a multiple of 10 times earnings—less than half the price-earnings ratio of the S&P 500—their market capitalization would be about $220 billion. Through exercising the warrants, bringing Fannie and Freddie out of conservatorship with a capital standard that allows them to price their business on an economic basis, and then selling the shares from its warrant conversion, the Biden administration could capture a very large portion of that $220 billion potential value for itself for whatever purposes it wishes, including an affordable housing fund.
From both a policy and a financial standpoint, therefore, Fannie and Freddie are worth far more to the Biden administration as vibrant companies run by private management to the benefit of homebuyers than as zombie companies run by FHFA to the benefit of banks. But to unlock that value, someone in the administration is going to have to step up and call the leaders of the Financial Establishment and the Federalist Society on their fictions about the two companies, and be willing to voluntarily cancel the net worth sweep without a judicial ruling saying it must do so, which will not be forthcoming. And the clock is ticking. In two years the opportunity to make defining policy and financial choices with respect to Fannie and Freddie may pass to the next administration, or the next, on towards 2040 or beyond.