Jim Millstein: An Administrative Plan to Restructure and Reform the GSEs

March 29, 2016

“The housing system we have today is unhealthy and unsustainable, mortgage credit remains overly tight, taxpayers remain at risk, and the system lingers in a dysfunctional limbo” (Parrott et al. 2016). Why? Because in 2011, the executive branch decided to pass responsibility for GSE reform to Congress rather than use the powers given it under the Housing and Economic Reform Act of 2008 (HERA). In turn, major congressional reform efforts were based on “winding the current system down and starting largely from scratch” (Parrott et al. 2016).  Ultimately, these efforts foundered on the fear that a “white board” approach to remaking an $11 trillion credit market for single- and multifamily residences1 would do more harm than good. In Congress’s hands, Fannie Mae and Freddie Mac proved not only too big to fail but also too important to kill. Hence our present state of “dysfunctional limbo.”

Worse, in 2012, the Treasury Department and the Federal Housing Finance Agency (FHFA) agreed to amend the preferred stock purchase agreements (PSPAs) to require that all of Fannie and Freddie’s profits be paid in perpetuity to the federal government (the Third Amendment). As a result of this amendment, $246 billion of earnings have been siphoned out of the companies for use in the general fund, depriving the conservator of the means to resolve the GSEs’ undercapitalization, the fundamental problem that HERA directed the conservatorships to address. FHFA has since taken an indirect approach to solving this problem through so-called “risk-sharing” deals, but, as former Fannie chief financial officer Tim Howard recently noted, less risk has been shared than meets the eye.2

With legislative reform stalled and Treasury bound to the defense of its Third Amendment in court, we are left with the hope that a new administration will use HERA’s authority to fix the structural flaws in the GSEs that the crisis revealed. Most of this can be done administratively by FHFA. All it takes is political will.

First, FHFA as conservator needs to do what Congress expressly mandated it to do under HERA: “ensure that each regulated entity operates in a safe and sound manner, including maintenance of adequate capital.”3 To fulfill this mandate, until Treasury is willing to amend the PSPAs, FHFA should suspend payment of cash dividends on Treasury’s senior preferred stock to ensure that the GSEs have the resources to build adequate levels of capital to protect taxpayers against loss on the GSE’s trillions of dollars of mortgage guarantees. FHFA Director Melvin L.Watt recently stated that he “expects Freddie Mac and Fannie Mae to determine their pricing as though they were holding capital and seeking an appropriate economic return on this capital” (Watt 2016). This exercise should not be theoretical. FHFA should allow the GSEs to adjust pricing and build capital to levels adequate for the expected losses on their guaranty liabilities.

Second, in anticipation of corporate restructuring (described below), FHFA should promulgate new regulations to fix the structural flaws that the crisis revealed in Fannie and Freddie’s corporate governance. There are three areas in which regulations need to be developed for the mortgage guaranty businesses: a new risk-based capital regime for different mortgage products, limits on leverage and on permitted equity rates of return, and business activity limitations.

Because the GSEs are private companies with public charters, there is an inherent conflict of interest between the GSEs’ obligation to promote access and affordability and the private market’s imperative to maximize shareholder value. Profit maximization clearly became paramount before the recent financial crisis. To retain market share and boost income, the GSEs chased an explosive market for private-label securities (PLS). They relaxed underwriting standards for conforming loans and used government-subsidized funding to purchase higher-yielding—and ultimately extremely risky—PLS. Although some of that activity expanded access and affordability, the imperative to maximize shareholder returns blinded managers to the credit risks they were taking. Weak capital regulation left them undercapitalized for the size of the levered portfolios and mortgage guarantees that their regulator permitted them to carry into the crisis. Meanwhile, the benefits of the funding subsidy from their implied government guaranty were pocketed at least as much by the companies’ shareholders as the homeowners whose mortgages they guaranteed, transforming a subsidy for affordable housing into an invidious welfare program for corporate shareholders.

These structural flaws can be remedied by new regulations well within the authority conferred on FHFA under HERA. First, to address the conflict between their affordability mission and shareholder returns, the mortgage guaranty businesses should be subjected to utility-like regulation, with strict activity, leverage, and return on equity limitations. Limits on leverage and on the earned rate of return on capital should prevent the GSEs from repeating many of the mistakes of the past crisis. Second, with new statistical information from the crisis on loss frequency and severity across different classes of mortgage products, a more nuanced capital regime can be developed and refined for the guaranty businesses, similar to what federal banking regulators have developed for banks. Enforcing new capital requirements and strict limits on leverage and equity rates of return should prevent managers from abusing their public charters for their shareholders’ benefit and from taking on outsized risks. Third, the federal backstop for the GSEs’ mortgage guaranty businesses should be made explicit and priced to reflect the government’s risk of providing it and its cost of doing so.

This last “reform” will require congressional action. But the FHFA—with Treasury’s consent under the PSPAs—can and should start the process now by restructuring the assets and liabilities of the GSEs to separate the explicit government guaranty the GSEs currently enjoy through the PSPAs from the mortgage guaranty businesses and then regulate by contract the pricing of the guaranty and the types of mortgage products to which it could attach. This administrative phase of reform would work as follows.

First, FHFA would cause each GSE to create a new wholly-owned subsidiary (Newco), transfering the assets of its mortgage guaranty business to Newco and having Newco assume the liabilities of each business. This could be done separately for the single-family and multifamily guarantee businesses. With this asset and liability transfer, Newco and the GSE would enter into a reinsurance contract, pursuant to which the GSE parent would “reinsure” Newco against any loss on its outstanding and future mortgage guarantees in exchange for a portion of the subsidiary’s guaranty fees (a “reinsurance fee”). The GSE’s liability for reinsurance to Newco would continue to be backstopped by Treasury’s outstanding commitment to purchase additional preferred shares under the PSPAs to cover the GSE’s losses. However, to protect itself against future calls on its backstop, Treasury could require that FHFA cause each GSE to use its reinsurance fees to build a mortgage insurance reserve fund (MIF) up to a certain level before “sweeping” any excess to Treasury as a dividend on its outstanding preferred stock.

Second, to protect the GSE against being called on its reinsurance of Newco’s guaranty liabilities (and to protect Treasury against being called on its backstop of that reinsurance), the reinsurance contract between the GSE and Newco would require Newco to use its portion of guaranty fees—after operating expenses and reserves—to build capital to absorb potential future losses on its mortgage guaranty liabilities. The actual level of capital to be retained over time would be consistent with risk-based capital regulations for the mortgage guaranty businesses to be developed by the FHFA under HERA. Once each Newco has sufficient capital to be deemed “well capitalized” under FHFA’s new capital regulations, FHFA would require that the reinsurance contract be amended so that reinsurance would not attach to future Newco guarantees unless Newco is in compliance with FHFA’s new business activity, leverage, and return on equity capital regulations.  Thereafter, FHFA would cause the GSE to sell its equity interests in each “well capitalized” Newco in a series of public offerings until all shares had been sold, thereby privatizing the GSE’s mortgage insurance businesses.  The proceeds from those stock sales and the proceeds of the ongoing liquidation of the GSEs’ portfolios would be applied as provided in HERA, first to the GSEs’ outstanding third-party debt obligations, and then to Treasury’s senior preferred stock outstanding under the PSPAs, and then to each GSEs’ other outstanding equity securities (including Treasury’s warrants) in order of liquidation preference.

The result of this corporate restructuring will be four new private, well-capitalized, well-regulated “first-loss” insurers standing in front of the GSEs on their entire books of mortgage-backed securities (MBS), plus a new fully funded MIF at each GSE standing in front of Treasury on its backstop of the GSEs’ solvency under the PSPAs. The GSEs, their charters, their MIFs, and their obligations under the reinsurance contracts with the newly privatized Newcos would remain in conservatorship backstopped by the PSPAs. MBS composed of “conforming” mortgages would be guaranteed by a combination of first-loss insurance provided by the Newcos in front of catastrophic risk insurance provided by the GSEs. The GSEs would continue to wrap the entire security so that the to-be-announced market continues to function. The reinsurance fees paid to the GSEs would fund the MIFs, pay for the operating expenses of the common securitization platform, fund the federal accounts created under HERA dedicated to affordable housing, and pay Treasury for its ongoing capital commitment under the PSPAs.

This would balance facilitating access and affordability with recapitalizing an undercapitalized system, tapping private investor demand for mortgage credit risk through the sale of the four Newcos’ equity to the public, and protecting taxpayers against future losses by building layers of new capital at the Newcos and MIF reserves at the GSEs. These administrative actions, consistent with HERA’s mandates, can charter a path to ending the longest-running conservatorships in American history without threatening the stability of the housing market or mortgage credit formation.

These administrative actions will also help Congress focus on the big picture: What support should the government provide, and how should the government regulate the conforming mortgage market? For example, the government guarantee currently extended by Treasury through the PSPAs to the GSEs could be restructured to operate outside of the conservatorships. One way would be to redefine the charter for one of the GSEs to reflect the utility-like model described above. That entity could either be sold out of conservatorship to private investors or be converted into a government corporation, as recently suggested by Parrott and colleagues (2016). Either way, it would have an explicit line of credit from Treasury for which it would pay a commitment fee. It would be regulated by FHFA with an appropriate risk-based capital regime and total leverage limitation. It would continue to have access and affordability goals. And it would continue to provide securitization services in conjunction with the common securitization platform and catastrophic risk insurance for an appropriate fee to private first-loss insurers, including the Newcos and other new entrants. The other GSE could either be put into receivership and liquidated or merged into the other government corporation, or sold out of conservatorship if benefits from competition in a tightly regulated utility model would be material.

Creating a government- or highly regulated private corporation through which to funnel the government guaranty to the conforming mortgage market is not too far from the Federal Mortgage Insurance Corporation proposed by Senators Johnson and Crapo in bipartisan legislation that stalled in 2014. With such legislation, FHFA could also license and regulate private first-loss insurers directly, replacing the indirect influence exercised through the GSE-Newco reinsurance contracts effective during the conservatorships.

Alternatively, Congress could eliminate the government guarantee for the conforming mortgage market altogether. In that case, after provisioning for the GSEs’ exposure on outstanding MBS, each GSE could be put into receivership and run off. Charters would be terminated. The grand experiment proposed by the Protecting Americans from Tax Hikes Act of 2015 could then be run with taxpayers out of harm’s way.

The structure proposed in this essay would not alter the basic system used by the government today to facilitate access and affordability in housing. The 30-year, prepayable, conforming mortgage would survive. There would be no disruption in the MBS market and the funding it provides homebuyers. In the first phase, the GSE charters would survive and be imposed through reinsurance contracts on the newly privatized mortgage guaranty businesses during the remainder of the conservatorships. Various programs administered by the departments of Housing and Urban Development and Agriculture would continue to deliver federal support for housing and mortgage finance apart from the market for conforming loans.

Access should be improved in two ways. First, private first-loss insurers facing a risk-based capital regime can facilitate demand for riskier mortgage credit while holding an adequate buffer for that additional risk. To date, the FHFA has been reluctant to allow the GSEs to take on riskier credit given their exposure to expected losses, exposure that this proposal would shift to private first-loss insurers. Second, the proposed system would provide resources to fund the Housing Trust Fund, Capital Magnet Fund, and the HOPE Reserve Account.

Appropriate capital requirements and private rates of return mean that the price of a conforming loan will  increase, all other things being equal. Guaranty fees charged by first-loss insurers can be regulated to some extent by the GSEs, whose necessary reinsurance provides significant bargaining power. But profits are necessary for private enterprise. There will be tradeoffs among affordability, access, private risk taking, and taxpayer protection.


  1. Mortgage Debt Outstanding,” Board of Governors of the Federal Reserve System, last updated March 11, 2016.
  2. Timothy Howard, “Risk Sharing, or Not,” Howard on Mortgage Finance (blog), March 9, 2016.
  3. Housing and Economic Recovery Act of 2008, Pub. L. No. 110-289 (2008).


Parrott, Jim, Lew Ranieri, Gene Sperling, Mark Zandi, and Barry Zigas. 2016. “A More Promising Road to GSE Reform.” New York: Moody’s Analytics. 

Watt, Melvin L. 2016. Prepared remarks at the Bipartisan Policy Center, Washington, DC, February 18. 

Jim Millstein is the founder and chief executive officer of Millstein & Co., a financial advisory firm with 30 professionals with offices in Washington, DC, and New York City. He is also an adjunct professor of law at Georgetown University Law Center, where he teaches federal regulation of financial institutions. He is also a commissioner on the American Bankruptcy Institute’s Commission to Study Reform of Chapter 11. 

From 2009 to 2011, Millstein was the chief restructuring officer at the US Department of the Treasury, where he oversaw and managed the department's largest investments in the financial sector and was the principal architect of AIG’s restructuring and recapitalization. From 2000 to 2008, Millstein was managing director and global cohead of corporate restructuring at Lazard. Before that, Millstein was partner and head of the corporate restructuring practice at Cleary Gottlieb Steen & Hamilton.

Millstein graduated summa cum laude with a BA in politics from Princeton University and graduated summa cum laude with an MA in political science from the University of California, Berkeley. He received a JD from Columbia Law School, where he was a Harlan Fiske Stone scholar. 

Neither Millstein nor Millstein & Co. holds any Fannie Mae or Freddie Mac securities or earns any revenue from Fannie Mae or Freddie Mac or from any investor in their securities.

Millstein is also a member of the Urban Institute's Housing Finance Policy Center's Council.