The Rescue of Fannie Mae and Freddie Mac-Module B: Senior Preferred Stock Purchase Agreements

The government passed HERA to enhance resolution and funding alternatives with respect to the GSEs in light of their severely weakened condition, their critical role in the stability of the mortgage and housing markets, and the stresses then impacting those markets. Prior to the passage of HERA, the OFHEO (the former GSE regulator) had no viable way to fund a conservatorship intervention, which rendered its conservatorship authority ineffective. HERA created a new regulator, the FHFA, and provided it with expanded authority over the GSEs. The new law also allowed Treasury to fund Fannie Mae and Freddie Mac for an emergency period (expiring December 31, 2009), thus ensuring their solvency. These emergency powers enabled Treasury to enter into the SPSPA, the MBS purchase program, and the credit facility. See Vergara (2021) and Zanger-Tishler and Wiggins (2021) for more information on the MBS purchase program and credit facility, respectively.

On September 7, 2008, the Treasury and FHFA announced a four-part intervention to stabilize the GSEs and maintain their solvency, consisting of (1) taking the firms into conservatorships, (2) entering into the SPSPA, (3) establishing the Credit Facility, and (4) launching the GSE MBS purchase program. Each component of the rescue plan was designed to address a particular set of constraints confronting the two GSEs. The SPSPAs were central to maintaining the solvency of the two firms, as it assured their positive net worth even with the likelihood of continued losses.

As stated in HERA, Treasury’s emergency powers included “purchas[ing] any obligations and other securities issued by [Fannie Mae and Freddie Mac]” (Housing and Economic Recovery Act 2008). These powers were unlimited as to amount but expired on December 31, 2009. The government could have funded the GSEs through continued purchases of debt or securities, similar to Treasury’s GSE MBS program, which purchased $225 billion of agency MBS ($220.8 billion face value) before expiring on December 31, 2009. (See Zanger-Tishler and Wiggins 2021). Yet no matter how large any such purchases might have been, they also would have had to cease as of December 31, 2009, leaving the GSEs with an uncertain future.

The government sought to avoid this result, but no one could predict just how long the conservatorships would last or how much funding would be needed. Therefore, the FHFA and Treasury aimed to structure a funding mechanism that complied with HERA’s limitations and had maximum duration so that it would provide for the GSEs’ funding needs after December 31, 2009, should the conservatorships last that long (Jester et al. 2018).

Designing the SPSPAs in the form of a “keep well”—an agreement typically entered into by a parent company wanting to guarantee the solvency of a subsidiary—allowed the FHFA and Treasury to fulfill these criteria. Instead of requiring continued purchases of any obligations or securities, the keep well agreement provided for a onetime sale of 1 million shares of SPS—with an initial Liquidation Preference of $1,000 per share—from each GSE to Treasury.^FIt is worth noting that during the crisis, the government often received preferred shares for its capital injections, for example, American International Group and the bank investments under TARP. In return, Treasury agreed to fund the GSEs through draws, with each draw increasing the aggregate Liquidation Preference, which would have to be paid to Treasury under certain circumstances.

The variable Liquidation Preference of the SPS and its accompanying dividend formula were central to the SPSPA’s funding mechanism. In return for this commitment (among other considerations), the GSEs paid Treasury a 10% dividend (12% if delinquent) based on the previous quarter’s Liquidation Preference through 2012 (Fannie 2008a). Additional draws increased the Liquidation Preference, and the Treasury’s investment in the GSEs’ senior preferred stock. This avoided the need for additional purchases. By relying on an increasing Liquidity Preference, the SPSPAs facilitated continued funding of the GSEs and maintenance of their solvency even after December 31, 2009, without running afoul of HERA’s limitations (Jester et al. 2018).

Despite the absence of an explicit government guarantee, GSE MBS and debt had historically been viewed as safe based on an implied guarantee, that is, the assumption that the US government would back these securities in the event of a crisis (FCIC 2011, 316). As a result, and because of the size and global presence of the GSEs, it was thought that the failure to honor these obligations, in particular senior debt, could undermine investor confidence and even trigger a global credit crisis (Jester et al. 2018).

Setting the SPS senior to all other stock but junior to all debt achieved the goal of protecting the GSEs’ creditors. Frame et al., among others, however, have questioned the government’s decision to protect not only senior debt but also subordinated debt. They suggest that haircutting these creditors, for example, might have promoted good market discipline, “signaling that [subordinated] debt is indeed risky” and “curbing moral hazard in similar institutions going forward” (Frame et al. 2015).  

When asked about this in an interview with the Yale Program on Financial Stability, former OFHEO Director James Lockhart, III suggested that the government knowingly accepted the risk of perpetuating moral hazard in light of the far greater risks involved with wiping out these creditors (Lockhart 2018). Treasury and FHFA legal counsel were convinced that imposing losses on junior creditors, which would have been consistent with market discipline, might not be possible in isolation. More specifically, they advised that doing so might also trigger “a cross-default on senior debt and [potentially even GSE] mortgage-backed securities.” Rescuing the GSEs on such terms would have risked triggering the exact kind of systemic crisis that the government intervened to prevent (Lockhart 2018).

Despite the perceived unfeasibility of imposing losses on creditors, the government saw the need to provide some level of protection for the taxpayers’ investment. That being the case, the GSEs placed the SPS senior to all other common and preferred stock. Shareholders hotly contested this arrangement, as it caused them to lose almost all of the value of their investment. In anticipation of such losses, Treasury asked banking regulators to look into potential consequences for banks holding large quantities of preferred shares. Banking regulators conducted a review and concluded that no systemic consequences would arise from losses borne by these banks (Lockhart 2018). A number of smaller banks, however, were significantly affected, and 15 failures and two distressed mergers of small banks either directly or indirectly resulted from the takeover (Lockhart 2018; Rice and Rose 2016).

Jester et al., who took part in designing the SPSPA, say the dividend rate was chosen “based on a review of the market for similar securities” (Jester et al. 2018). If a GSE failed to pay a dividend on time, the rate reset to 12% until all overdue dividends were paid in full.

The FHFA identified two factors in its decision to implement a variable dividend. For one, this approach, which provided the GSEs and Treasury with more flexibility, eliminated the need for the GSEs to borrow from Treasury to pay dividends, a practice that increased the Liquidation Preference, and by extension, future dividend payments under the SPSPAs (FHFA 2012). FHFA director Edward DeMarco argued that the new approach increased market confidence (FHFA 2012). Secondly, the FHFA asserted that Fannie Mae’s and Freddie Mac’s earnings would be “used to benefit taxpayers,” as taxpayer money had sustained the GSEs (FHFA 2012).

This new formula was not without risk to Treasury; dividends became dependent on the GSEs being profitable. If the GSEs did not realize profits, they did not pay the dividend. This was the case for Freddie Mac in fourth quarter 2015 and second quarter 2016 (see Appendix B). 

The variable dividend was also part of a larger initiative that aimed to build a new infrastructure for the secondary mortgage market in contemplation of a significant shrinking of the GSEs’ operations and role. An additional significant factor was the required reduction of the GSEs’ capital reserves (Thompson and Wiggins 2021).

In addition to the purchase of SPS, Treasury also received under the SPSPAs (among other considerations) a warrant to purchase up to 79.9% of the common stock of each GSE for a nominal amount at any point during the warrant’s 20-year term, offering taxpayers significant upside in the event the GSEs were restored to viability (Treasury 2008a).

A maximum of 79.9% of all shares was chosen in consideration of an accounting rule that would require the federal government to take the GSEs onto its balance sheet if it acquired a stake in them of 80% or more. Jester et al. (2018) have insisted that, absent this rule, the authors, who served as financial regulators and designed the SPSPAs during the crisis, would have sought to give taxpayers full ownership of the GSEs. The authors opted for warrants to purchase only 79.9% of all shares, however, because they believed that the risk of “adding trillions of GSE debt and guarantees to the federal balance sheet” far outweighed any potential benefit of owning an additional 20% of shares (Jester et al. 2018).

Given the absence of consensus on the GSE’s funding needs, Treasury had asked for and received under HERA authority to fund the GSEs in an unlimited amount through December 31, 2009, subject only to the federal debt ceiling, which the law had raised by $800 billion to accommodate the rescue (HERA Section 3083, Jester et al. 2018, 4, 9). In designing the SPSPA, the Treasury sought to provide funding for each GSE that was large enough to “eliminate any reasonable concern.” However, there was concern about setting the amount too high (doing so might signal that the problems were larger than they really were), and legal counsel advised against making an unlimited commitment (Jester et. al. 2018, 9-10).Thus, officials settled on $100 billion per GSE, as this amount exceeded most market projections of their losses at the time (Jester et al. 2018, 9-10). Even if $100 billion per GSE proved to be inadequate, Treasury believed the incoming administration could amend the SPSPAs to raise the cap, if needed (Jester et al. 2018, 9-10).

Defaults on subprime and Alt-A mortgages in Fannie Mae’s and Freddie Mac’s mortgage portfolios had contributed to the GSEs’ losses during the crisis (FCIC 2011). The SPSPAs initially limited the portfolio cap for each GSE to $850 billion, and later raised it $900 billion, because officials recognized that the GSEs needed to continue purchasing loans in order to keep the secondary mortgage market from collapsing (Treasury 2008b). Secretary of the Treasury Henry M. Paulson, Jr., argued that the portfolio cap would “promote stability in the secondary mortgage market and lower the cost of funding, [since] the GSEs will modestly increase their MBS portfolios through the end of 2009” (Treasury 2008b). After the GSEs had supported the market through the most severe period of the crisis, the SPSPAs were amended to require the GSEs to begin to wind down their portfolios to reduce systemic risk (Treasury 2008b). Paulson commented on this change—“In 2010 [GSE] portfolios will begin to be gradually reduced at the rate of 10 percent per year, largely through natural run off, eventually stabilizing at a lower, less risky size” (Treasury 2008b).

Under the SPSPA, the GSEs could not increase their debt levels to more than 110% of their debt on June 30, 2008. The first amendment, passed on May 6, 2009, raised the debt limit to 120% of the portfolio cap. Operating in conjunction with a mandatory decrease of the portfolio, the debt cap decreased by 10% annually beginning in 2010 (UST/Fannie 2009a). The third amendment accelerated the decrease from 10% to 15% (UST/Fannie 2012).

Treasury and the FHFA recognized that the GSEs required adequate leverage to purchase and hold mortgages in their portfolios. Since the GSEs were obligated to reduce their portfolios, the FHFA limited outstanding debt as a percentage of the GSE’s portfolio.

Without Treasury’s approval, the GSEs and the FHFA, as conservator, could not:

1. Pay dividends on any class of stock other than the Senior Preferred Stock,
2. Issue common or preferred stock,
3. Enter into a contract with any affiliate of Treasury,
4. End the conservatorship, unless the FHFA moved the GSEs into receivership,
5. Sell, convey, or transfer any of its assets, unless the transaction constituted “the ordinary course of business,”
6. Merge with another entity,
7. Avoid compliance with SEC deadlines for 10-K, 8-K, and other reports, or
8. Enter into new executive compensation arrangements (UST/Fannie 2009a).

The third amendment also capped the GSEs capital reserves at $3 billion (beginning in 2013) and required reductions of $600 million annually until they reached zero in January 2018. At that point, the GSEs would depend solely on Treasury funds to cover any losses (UST/Fannie 2012). The reduction in GSE capital reserves in 2012 seemed to align with the Obama administration’s interest in shrinking the GSEs’ operations and their role in the secondary mortgage market but also might have complicated any plan to have them exit from conservatorship (Thompson and Wiggins 2021).

By 2017, the administration had changed, and the FHFA and GSEs sought amendments to the SPSPA. By letter agreement dated December 21, 2017, Treasury and the FHFA agreed to raise the maximum capital reserve amount from zero to $3 billion indefinitely, beginning in January 2018 (UST/Fannie 2017).The amendment was intended to permit the GSEs “to maintain a limited capital buffer in an amount that should be sufficient to cover income fluctuations in the normal course of business” (UST/Fannie 2017). It is unlikely that the change was instituted to help the GSEs prepare for exiting the conservatorship, since the letter agreement mandated that the GSEs also increase the Liquidation Preference of their respective SPS by $3 billion, which the firms have to pay down before exiting the SPSPAs (UST/Fannie 2017). Treasury also asserted that the GSEs could maintain a $3 billion capital buffer only for as long as they continued to pay dividends (UST/Fannie 2017). (Also see “Outcomes” for discussion of later amendments to the SPSPA.)

Under the SPSPA, the GSEs were required to pay a quarterly Periodic Commitment Fee, accruing from January 1, 2010, and payable quarterly beginning March 31, 2010, to compensate Treasury for its ongoing commitment (UST/Fannie 2009a). The FHFA and Treasury, in consultation with the Federal Reserve, were to decide the amount of the Periodic Commitment Fee, which would be applicable for a five-year period (and reset every five years) (UST/Fannie 2009a). Treasury could, in its discretion, waive the Periodic Commitment Fee in 2010 and at the beginning of each subsequent quarter based on adverse conditions in the mortgage market (UST/Fannie 2009a). The GSE could, at its discretion, choose not to pay the Periodic Commitment Fee but have it added to the aggregate Liquidation Preference (UST/Fannie 2009a). Since the GSEs continued to post losses into 2012, Treasury waived the commitment fee for each quarter beginning in March 2010 until it suspended the fee with the third amendment (UST/Fannie 2012).

The FHFA and Treasury circulated press releases and detailed reports as part of their transparency policies. Treasury and the FHFA announced amendments to the SPSA when agreed to even though they might take months to become effective, allowing the market to adjust to the announced changes before they were instituted. For example, most components of the third amendment—executed in August 2012—did not take effect until first quarter 2013. Treasury and the FHFA also made copies of each amendment easily accessible via their websites (FHFA 2021a).

Treasury and FHFA officials underscored the importance of protecting taxpayers in their reports and public statements. During Treasury’s announcement of the GSE intervention, Treasury and Secretary Paulson stressed that protecting the taxpayer remained one of Treasury’s chief priorities (Treasury 2008a; 2008b). In February 2010, Edward DeMarco, then acting director of the FHFA, wrote a letter to Congress acknowledging that taxpayer funds kept Fannie Mae and Freddie Mac operating (DeMarco 2010). DeMarco also pledged to outline how the FHFA had limited, and would limit, GSE losses (DeMarco 2010). DeMarco later argued that the third amendment was enacted mostly to protect taxpayers (FHFA 2012).