Notes from the Vault

Larry D. Wall
May 2014

Have U.S. taxpayers been (almost) fully repaid by Fannie Mae and Freddie Mac? I raised this issue in the March Notes from the Vault, contending that the two government-sponsored enterprises (GSEs) had not fully repaid the Treasury for the value of its assistance since they were placed into conservatorship in September 2008. Richard Epstein, the Peter and Kirsten Bedford Senior Fellow at the Hoover Institution, responded to my post with the argument that, regardless of how one views the economics starting in 2008, the relevant point for conducting the analysis is the third amendment to the GSEs' senior preferred stock purchase agreements signed in August 2012. That agreement was significant because it required Fannie and Freddie to pay all of their profits through 2017 to the Treasury. Epstein argues that by August 2012, the financial condition of the GSEs had improved to the point where the Treasury's claim was limited to its existing holdings of senior preferred stock.

This month I focus on the economic role of taxpayer support of Fannie and Freddie since August 2012. My discussion addresses four issues: (1) the content of the third amendment, (2) Epstein's critique, (3) the condition of the GSEs in August 2012, and (4) alternative actions the Treasury could have taken in 2014.

Third amendment to the GSEs' senior preferred stock purchase agreements
At the time of the third amendment, the U.S. Treasury had purchased $187.5 billion in senior preferred stock to prevent the GSEs from becoming insolvent. The GSEs were paying a 10 percent per year dividend on this stock, with the payment of dividends through 2011 funded by additional Treasury purchases of senior preferred stock.

Although the third amendment made several changes to the original agreement, two changes are especially noteworthy. One part that Epstein (and others) pointed to is the change to the required dividend payments that obligated the GSEs to pay all of their future profits to the Treasury as dividends. This amendment also implied that the GSEs would remain in conservatorship through at least 2017.

Another important change is that the periodic commitment fee was suspended so long as the third amendment's dividend payment requirements remained in effect. The original agreement had included provisions for a periodic committee fee "to fully compensate Purchaser for the support provided by the ongoing Commitment following December 31, 2009." The initial agreement also provided that the commitment fee would be "mutually agreed by Purchaser and Seller, subject to their reasonable discretion." However, the purchaser (Treasury) "may waive the Periodic Commitment Fee for up to one year at a time, in its sole discretion, based on adverse conditions in the United States mortgage market." The Federal Housing Finance Agency (FHFA) Office of Inspector General says that the fee had never been set and the Treasury consistently waived this fee prior to the third amendment.

Professor Epstein's analysis
Many articles appeared earlier this year noting that Fannie and Freddie were about to have paid dividends in excess of the $188 billion invested by Treasury in the GSEs' senior preferred stock. In a previous commentary, Professor Epstein provided a clear statement of the message that was implicit in most of these stories: Once the $188 billion was paid to the Treasury, taxpayers would be fully compensated and the GSEs should be returned to their private shareholders.

My earlier Notes from the Vault made four points in response to this argument. First, this method of accounting would deny the dividends that the Treasury should have received on its senior preferred holdings in 2012–14. My back-of-the-envelope calculation showed that about $18 billion would have been owed in 2012 alone. Second, this analysis ignores the value of the Treasury's commitment to bear losses in excess of $188 billion on the GSEs pre-conservatorship book of business. Third, the 10 percent rate paid on the senior preferred stock likely undercompensated the Treasury, given that junk bond yields at the time were over 15 percent. Finally, Fannie and Freddie had only been able to book profitable new business since being placed into conservatorship because of the Treasury's guarantees. Most of the post-conservatorship "profits" that the GSEs' private shareholders were claiming should be viewed as a return to taxpayers for bearing all of the risks associated with the GSEs' new business.

Professor Epstein concurred that the $188 billion figure mistakenly excluded 2012–14 dividends. However, he notes that this argument does not dispute his point that the GSEs have almost fully repaid the Treasury, as it merely adds a few billion more dollars onto the required payments. I agree that this argument, by itself, only shifts the point at which the Treasury would have been repaid.

Professor Epstein implicitly combines my two arguments about the guarantees on the pre-conservatorship and post-conservatorship books of business and offers a rebuttal. He acknowledges that the guarantees had some value at the time the GSEs were put into conservatorship, but he notes that this value was calculable (a point I also made). He continues then with his main argument:

By the time the Third Amendment came about, no additional commitments would be needed, so that these contingent liabilities were not a serious factor in figuring out whether the Third Amendment was fair to the shareholders—which given its wholly one-sided nature it was not. (emphasis added)

This response neither accepts nor denies my argument that the Treasury (and by extension, the taxpayers) deserved compensation for bearing the contingent risk. Instead, it argues that since the Treasury has not committed any additional funds since 2012, that is sufficient to prove that the value of the Treasury's contingent liabilities was not material and hence could not justify the terms of the third amendment.

Epstein's third argument is that the Treasury also received the option to purchase 79.9 percent of the GSEs' stock at a nominal price and that this compensated for the low dividend rate. Moreover, he argues that the relevant time to value these options was in 2012 when the third amendment was adopted. However, even at the time of the third amendment the taxpayers were still bearing essentially all of the risk, as we shall see below. Why should taxpayers be satisfied with 79.9 percent of the GSEs' value if the GSEs fully recover when taxpayers bear 100 percent of the additional losses if the GSEs do not?

Thus, while we have differing views about conditions in 2008, Epstein's primary argument is that the GSEs were in much better condition in 2012.

The GSEs' financial condition in 2012
The third agreement was signed in August 2012, so I will focus on the third-quarter financial statements (dated September 2012) of Fannie and Freddie.1 The third-quarter 2012 Conservator's Report shows that the GSEs had accumulated deficits in retained earnings that were exactly offset by the Treasury's holdings of $188 billion in senior preferred stock. During the third quarter, the GSEs reported sufficient earnings to pay the required dividends on Treasury's senior preferred stock and increase their combined capital from $3 billion to $7 billion. Since then, both Fannie and Freddie have continued to report earnings in excess of the dividend payments that would have been required on the senior preferred stock. However, the GSEs have not been able to retain earnings to rebuild capital under the net sweep agreement, as any buildup in equity was payable to the Treasury.

Given the GSEs' return to profitability and $7 billion in capital, were they in such a strong financial position that the continuing Treasury guarantee was of little value? The two GSEs have been profitable since 2012, but conditions have also been relatively favorable with a growing economy, generally increasing real estate prices, and stable interest rates. However, the purchasers of mortgage-backed securities (MBS) guaranteed by the GSEs and the investors buying other GSE debt at tiny spreads over Treasury are not doing so merely because they expect the GSEs to be able to honor their commitments in favorable economic conditions. The GSEs' MBS guarantees and other GSE debt are highly valued because investors believe the GSEs' promises will also be fully honored in bad economic conditions. As continued good economic conditions were not guaranteed in 2012, the question of whether the GSEs had the financial strength to honor their promises in more trying economic conditions is essential to understanding whether their business model was viable without the Treasury's guarantee.

One of the primary measures of a financial firm's ability to absorb losses is a measure of its owner's equity relative to some measure of risk exposure, its capital adequacy. If the GSEs were in a strong financial condition such that the Treasury guarantee was unimportant, they should certainly exceed the minimum capital levels set by 12 U.S.C. 4612(a) which is the sum of: (1) 2.5 percent of aggregate on-balance sheet assets and (2) 0.45 percent of the unpaid balance of outstanding MBS issued or guaranteed by the GSEs (and other off-balance sheet obligations). Moreover, this is truly a minimum, as both GSEs exceeded these requirements prior to the crisis but still needed government support.

Fannie reported over $3.2 trillion in total assets as of September 30, 2012, and Freddie reported over $2 trillion in assets in the same time frame. Even ignoring the capital required to back the GSEs' guarantees, the combined GSEs had a capital-to-asset ratio of under 0.15 percent. Not only is this level far below the minimum capital level set by law, it is far below the critical capital level set in 12 USC 4613(a) of 1.25 percent of assets (plus additional capital for outstanding MBS guarantees and other off-balance sheet items). The significance of falling below the critical capital level is that this would imply the GSEs would be defined as "critically undercapitalized" under 4 12 U.S.C. 4614(a)(4) and subject to being put into receivership under 12 U.S.C. 4617(a)(3)(K).

The two GSEs were not (and currently are not) classified as critically undercapitalized only because the FHFA determined that "it is prudent and in the best interests of the market to suspend capital classifications of Fannie Mae and Freddie Mac during the conservatorship, in light of the United States Treasury's Senior Preferred Stock Purchase Agreement."

Alternatives to the third amendment
The Treasury had at least three other options available at the time of the third amendment in August 2012:

  1. The FHFA could have started applying the capital classifications to the GSEs, found them to be critically undercapitalized, and put Fannie and Freddie into receivership.

  2. The Treasury could have stopped waiving the periodic commitment fee and started charging the GSEs a fee that would compensate taxpayers for the risk they were bearing.

  3. The Treasury could have continued to waive the commitment fee. In this case, the profits in excess of senior preferred stock dividends could either be immediately used to repurchase the senior preferred stock or allowed the GSEs to rebuild their retained earnings as a prelude to some future repurchase of that stock.

The Treasury was highly unlikely to put the GSEs into receivership until an acceptable set of substitutes for their role in mortgage finance was in place. So the real issue is, who should receive the profits in excess of required dividend payments?

While the shareholders in the GSEs certainly favor the continued waiving of the commitment fee with the intent of ultimately returning the GSEs to their private shareholders, I find it very difficult to make the case that this option would have been fair to taxpayers. At the time of the third amendment, the private shareholders were making at most a trivial contribution to the financial strength of Fannie and Freddie. Even if we attribute all $7 billion of the GSEs' capital in September 2012 to the private shareholders, that would give the GSEs a tiny capital-to-asset ratio that falls far short of the standard to avoid being classified as critically undercapitalized.2 Further, even if the supervisors had been willing to allow the GSEs to operate with such little capital, the GSEs' MBS guarantees would have been accorded far lower value by investors and new debt issued used to fund their portfolios would have had to pay substantially higher credit risk premiums.3

That leaves the third amendment's income sweep or the payment periodic commitment fees as the only options that would be fair for taxpayers. Periodic commitment fees certainly have had the appearance of being fairer to the remaining shareholders, as these were a part of the original senior preferred contract. Whether a profit sweep versus fair commitment fees (fees fair to taxpayers) would have resulted in significantly lower payments from the GSEs to the Treasury is less clear. One way of estimating fair fees would have been to price the commitment using sophisticated financial methods along the lines discussed in my earlier post. A simpler alternative for calculating the value of the Treasury's commitment would be to compare the relative contributions of the private shareholders and Treasury to the financial strength of Fannie and Freddie. This latter method would suggest that the periodic commitment fee should have been equal to (virtually) all of the GSEs' profits as the taxpayers were bearing (virtually) all of the risks in August 2012.4

Larry D. Wall is the executive director of the Center for Financial Innovation and Stability at the Atlanta Fed. The author thanks Scott Frame for helpful comments on the paper. The views expressed here are the author's and not necessarily those of the Federal Reserve Bank of Atlanta or the Federal Reserve System. If you wish to comment on this post, please e-mail


U.S. Department of the Treasury (2014). Department of the Treasury's Memorandum in support of its motion to dismiss the complaint or, in the alternative, to transfer or stay the action. Continental Western Insurance Company versus Federal Housing Finance Agency et al. United States District Court for the Southern District of Iowa Central Division, Case 4:14-cv-00042 (filed April 29).

1 If anything, the use of the third-quarter statements gives a stronger picture of the GSEs' financial condition than would the use of second-quarter 2012 statements.

2 That is, attribute all of the earnings to the private shareholders despite the fact that the profits the GSEs had earned were largely attributable to earnings from businesses that relied on continuing government guarantees.

3 That the GSEs could easily have needed to make additional draws on the Treasury is demonstrated by FHFA's projections of the enterprises' financial performance in October 2012. The three scenarios considered showed the need for additional draws of $3 billion to $22 billion. More recent analysis using an even more stressful scenario found larger draws could be necessary. The April 30, 2014, projections for the "DFAST Severely Adverse Scenario" found that the GSEs could need between $84 billion and $190 billion in additional capital.

4 The Department of Treasury (2014) points out one economically important difference between the profit sweep and commitment fees would arise if the GSEs had low or negative earnings. If the GSEs were unable to pay the senior preferred stock dividends and the commitment fees from current earnings, the shortage would have to be covered by an increase in a GSE's draw on the Treasury's commitment to purchase senior preferred stock. However, the Treasury's ability to purchase senior preferred stock is capped, and the Treasury no longer has the authority to increase this cap. Thus, compensating taxpayers for their risk bearing by higher commitment fees would increase the risk that the GSEs would sustain losses in excess of the Treasury's ability to provide support.