Housing Finance Part 3: What to do with the GSEs?

Shutterstock 593820728

What would happen to the multi-trillion dollar housing market if the GSEs were gone or very much changed? This paper is third in a three-part series explaining housing finance and the role the government-sponsored enterprises play in capital markets.

Few believe the current housing finance arrangement is sustainable. Last year, the federal government supported seven out of ten new mortgages and refinances.1 Fannie and Freddie exist today under government conservatorship. Investors’ willingness to take on mortgage debt without receiving a government guarantee remains at all-time lows—private label mortgage securities (those not receiving guarantees from the federal government) make up less than 1% of the housing market, down from 2006 record highs of 55%.2

Thus, reform proposals abound. Some preserve Fannie and Freddie, but in a reduced role. Others unwind Fannie and Freddie and shut them down. If GSE reform is to happen, it is likely to radically reshape the world of housing finance.

How different will that world be? Start with credit risk. The main role of Fannie and Freddie is to absorb mortgages’ credit risk to support affordable housing and the secondary mortgage market. When mortgages fail, the GSEs cover the costs associated with default. But in tomorrow’s world, we could instead call upon someone, some other entity, or some investor—who, today, is not taking on that risk—to take some of it on. Most agree the current level of credit risk subsidization from the government is too high, putting taxpayers at risk. But in this new world, no one concretely knows if investors will come flocking to take the risk on.

Let’s assume the best and that private demand for credit risk is bursting at the seams.3 The result will still mean higher interest rates for mortgages, because no private guarantee will be as cheap as today’s government guarantee. Said another way, if reforms are implemented, tomorrow’s mortgages will be more expensive, period.4 But maybe that’s a good thing? Some argue that such reforms would make for a heathier marketplace since credit would flow only to the creditworthy and, in turn, curb the chances of another catastrophic housing bubble.

Understanding these two points—who would take on credit risk, and how much it would cost mortgage holders—is key to understanding the goals of GSE reform proposals. Each one not only establishes a system to incentivize private capital, but also addresses the inevitable rate increases to keep mortgages affordable.

Finally, this debate could open up a larger discussion on housing. The infrastructure that makes up our country’s federal housing policy is the product of decades-long policy debates with conflicting goals. What we’re left with is a patchwork of agencies, regulators, and policies that may no longer make sense. Given the magnitude of GSE reform, this serves as a unique opportunity to ask those big picture questions in housing policy: Is the 30-year fixed-rate mortgage still viable and desirable? Are renters a bad thing for the economy? Should the government be in the business of subsidizing one type of homeowner over another?

I. Backdrop: Who Exactly Are Fannie and Freddie?

The United States government has played an active role in the mortgage market dating back to the Great Depression.5 Before its involvement, mortgage rates fluctuated from region-to-region. Most loans were for less than 10 years and had large balloon payments at the end, meaning homeowners had to string together loan after loan or else risk foreclosure.6 Fannie Mae and Freddie Mac were created to fix this. Even though each was established decades apart (Fannie in 1938 and Freddie in 1970), their formation each had the goal of establishing and maintaining a healthy secondary mortgage market. A vibrant secondary mortgage market is needed to mitigate the inherent problems large, illiquid assets, such as mortgage loans, carry with them.

For decades, the Fannie and Freddie housing finance system worked well—until disruptions in home prices pushed these two mortgage giants to the brink of collapse in 2008. Mere hours before total meltdown, the federal government stepped in, placing Fannie and Freddie into conservatorship. Since 2008, the GSEs have required $187 billion worth of backing from the Treasury. The terms of the conservatorship included a drastic change in vision for Fannie and Freddie. No longer were they in the business of maximizing profit for shareholders, but instead were in the business of reducing risk to the taxpayer.7 Both GSEs were ordered to halt all lobbying activities and fire their CEOs. Both Fannie and Freddie were forced to shrink their portfolios and turn over most of their profits to the Treasury. To date, the GSEs have paid nearly $225 billion to the Treasury.8

The starkest reality of this seven-year-long GSE conservatorship is that the federal government continues to act as an oversized giant in the world of housing finance. About 73% of all mortgage originations in the secondary market today involve some form of government guaranteed securities. Before the housing bubble, this number was around 50 percent.9 Most policymakers agree that there needs to be more private capital involved to take on credit risk and that the government cannot continue to keep subsidizing housing at current levels.

II. The Regulators: What’s Being Done Outside Congress?

Absent Congress, two regulators—the Federal Housing Finance Agency (FHFA) and Consumer Financial Protection Bureau (CFPB)—have acted aggressively under their statutory authority.

The Federal Housing Finance Agency’s Goals: “Maintain, Reduce, Build”

As Fannie and Freddie’s regulator, the FHFA has been active on what Director Mel Watt calls a policy of “Maintain, Reduce, and Build.”10

“Maintain” refers to a continuation of the GSEs’ original goal: to preserve a liquid, efficient, competitive, and resilient housing finance market. An example is FHFA’s clarifying of the system of representations and warranties (“reps and warrants”). When Fannie and Freddie purchase a loan from a lender, they need assurance that the supporting documents provided by the lender are correct. Should Fannie and Freddie uncover inaccuracies, like the understatement of a family’s monthly income, they reserve the right under reps and warrants to force a lender to repurchase the loan. From the original lenders’ perspective this is known as “put-back risk.” Poor underwriting standards led to a lot of put-backs from 2006-2008.11 The hangover from that period has generated substantial amounts of uncertainty around the reps and warrants process. Lenders have responded to this uncertainty by expanding loan requirements for borrowers (i.e. larger down payments, higher FICO scores to qualify for a loan, etc.). These requirements are layered on top of the standards set by Fannie and Freddie. These new “layers” have made mortgages more expensive and harder to get. By clarifying the reps and warrants terms, Director Watt and FHFA’s goal is to restore certainty to the market, get lenders to stop layering on these added requirements, and reduce the cost of mortgages.

“Reduce” means reducing risk to the taxpayer. The GSEs are now engaging in risk-sharing transactions in which private market investors purchase securities that include some of the credit risk Fannie and Freddie take on whenever they guarantee a mortgage-backed security (MBS). As of early 2015, Fannie has engaged in six such deals, named Connecticut Avenue Securities (CAS), laying off 11.4% of its book of business. Freddie has had 11 deals, named Structured Agency Credit Risk (STACR), laying off 16.1% of its total book of business.12 These deals are sort of like a trial run to see how much appetite there is in the private sector for Fannie and Freddie credit risk. Sharing this risk reduces taxpayer exposure to another bailout.

Lastly, the “build” component of Director Watt’s goals involves FHFA’s attempt to create a common securitization platform for the GSEs. Even though Fannie and Freddie issue securities that receive identical, implicit government guarantees, historically, the securities issued by Freddie have traded at a disadvantage to Fannie. This disadvantage is due in large part to the considerable size advantage Fannie has over Freddie. The result is a pricing imbalance between the two GSEs’ securities. A common platform should eliminate this imbalance since shared contractual and disclosure requirements will likely lead to deeper, more liquid markets for the GSE securities. As with any security, having deeper, more liquid markets helps price the security more accurately.

Consumer Financial Protection Bureau’s Goals: Better Mortgages and Servicing.

CFPB has taken several major actions that impact housing finance. The most significant are the Qualified Mortgage rule (QM) and its cousin, the Qualified Residential Mortgage rule (QRM). The goal is simple: blow up the “originate-to-distribute” model that contributed to the financial crisis. The process is this: QM created guidelines for lenders to make “gold standard-quality” loans. QRM made it so anyone creating mortgage securities with loans below this gold standard must put skin in the game to cover a share of any losses should mortgages default.13 However, this rule is a balancing act—on the one hand, it injects much needed accountability back into the system, while on the other, it makes it more difficult to get a mortgage. Finding that balance will be a continual challenge.

III. Imagine There’s No Fannie: Key Questions for Congress on the GSEs

Many advocacy organizations have written excellent side-by-side comparisons of the competing GSE reform proposals.14 Here, we pose a series of questions policymakers should ask when assessing these proposals.

How do we bring private capital back into housing? How will future government guarantees work? How will they be priced?

Today, the decision for how much private capital is needed to receive a government guarantee is driven by decisions made at FHFA and the Federal Housing Administration (FHA). If a mortgage has a high loan-to-value ratio (a metric that compares the size of a loan to the value of the home: typically anything above 80% is considered high) the agencies will require the borrower to purchase private mortgage insurance. Additionally, in order to receive a guarantee from a GSE, FHFA will set (and often adjust) the guarantee fees (“g-fees”). A g-fee is the amount charged to borrowers for a GSE guarantee. This guarantee protects investors in the case of default, similar to an insurance premium.

There remains a contentious disagreement among lawmakers about what level of support the government should provide in the housing markets. Should these guarantees even exist? Should there be an explicit government guarantee beyond those provided by FHA?

Some lawmakers feel that there should be an explicit government guarantee, but only after private, secondary market capital is put up—ranging from 5% to 10% of the MBS. In this new world, guarantors (maybe even Fannie and Freddie) would act as facilitators, aggregating private capital before an MBS is issued. This private capital would be in a first-loss position, meaning in the event where foreclosures grow so large MBS trusts can no longer pay investors, this private capital cushion would be the first to lose. Once that cushion is depleted, the government (likely a new government entity) would step in, guaranteeing that investors get on-time payments of their MBS. Many refer to this as the government providing a “catastrophic government guarantee.” According to analysis from BlackRock Solutions, the risk management division of the world’s largest asset management firm, the recent crisis would have been adequately covered had there been a 4% capitalization cushion.15

Other lawmakers feel that capital markets alone should bear the risk of mortgages. In the event of a catastrophe like in 2008, the government would stay out. However, such a proposal wouldn’t completely abandon government involvement in guarantees. Instead, this proposal would ask for a major regulatory overhaul to encourage the use of covered bonds—a tool that would supplement the GSE guarantee.

Covered bonds are a relatively unknown product in the U.S. but popular in Europe. They are very similar to MBS in that they are backed by assets. But, unlike MBS, they stay on the bank’s balance sheet and are designated as part of a “fenced-in” pool. This pool is left untouched and the bank instead is allowed to pay investors back using their general funds. The “fenced-in” pool is only used to repay bondholders in the event that the bank becomes insolvent. Supporters of this covered bond idea argue that it creates a framework that could essentially absorb mortgages’ credit risk and replace the need for a government guarantor like the GSEs.

Still unknown, however, is whether covered bonds can entice enough capital to support America’s enormous appetite for affordable mortgage credit.16 A Chicago Federal Reserve study found that covered bonds can increase liquidity and make it easier for lenders to modify loans that run into trouble.17 Supporters also cite the success of covered bonds in Europe as evidence that a government guarantee is obsolete.

It is difficult to project whether the success in Europe would transfer to the U.S. because of the unique makeup of the U.S. mortgage market. The only European country remotely comparable in reliance on long-term fixed rate mortgages is France. Most European homeowners share much more of the risk associated with mortgages by taking out adjustable-rate mortgages (interest rate risk) and having to pay prepayment penalties (prepayment risk).

Being Mindful of Leverage in Housing Finance

The concept of leverage is important to this discussion and should be considered carefully by policymakers. Leverage is a term used in finance to measure how much money is being borrowed in order to purchase something—in this case a security. This was a serious problem in the U.S. in the lead up to the crisis for many financial institutions.

Dodd-Frank tackles this problem for some, but not all financial institutions. Regulators can now set capital requirements and leverage ratios for insured depository institutions, bank holding companies, savings and loan holding companies, and non-bank financial companies supervised by the Federal Reserve. However, untouched were Fannie Mae and Freddie Mac—the guarantors of trillions of dollars of mortgages. In all likelihood, the drafters of Dodd-Frank left GSEs out from this calculation because they knew that in order to address leverage requirements at the GSEs, they needed to first decide what the new housing finance system would look like—which meant reforming the GSEs.

Some capital markets watchers believe that any reform that allows housing securities to be freely leveraged would again expose us to another liquidity meltdown similar to the ones we saw in 2008 and the S&L crisis prior to that. By placing a cap on the amount of leverage that can be taken against these securities—especially a leverage cap for the first-loss capital—investors would be required to keep more skin in the game, mitigating the risks our economy will likely face during the next down housing market.

What role should the government play in affordable housing and mortgages?

Few doubt the impact all reform proposals will have on mortgage rates—each will result in higher priced mortgages.18 This raises concerns that disadvantaged families will be excluded from homeownership.

There are two general approaches to address affordable housing. The most common approach is to apply fees to every covered security the government guarantees. Those fees would be designated to trusts meant exclusively to put disadvantaged families and first-time homebuyers into a home. This cross-subsidization approach means everyone purchasing MBS backed by the government would share the responsibility of supporting affordable housing programs.

The other approach involves no fees. Instead the government’s role in affordable housing would be centered on programs housed within agencies like Housing and Urban Development and the Department of Agriculture. The rationale here is that the best way to tackle affordable housing is through targeted government programs.

What to do with Fannie and Freddie?

Almost every major GSE proposal out there has provisions to stop Fannie Mae and Freddie Mac from issuing, guaranteeing, or purchasing any mortgage-backed security. However, policymakers should consider how reform packages deal with the trillions of dollars’ worth of securities that are currently on the GSEs’ books. A key question is how quickly these legacy securities would be unwound while the new framework is put in place.

The GSEs have already begun to engage in risk sharing transactions with the CAS and STACR deals. In order to wind the GSEs down, the reform proposal would need to continue these transactions. The question is how much appetite investors have for risk-sharing transactions. Can the entire book of business get unloaded onto the markets? Some argue that dragging this “unloading” out for years is the best solution because it will provide ample time for the markets to absorb this risk. Their proposal would provide more flexibility and an opportunity for the regulator to set the rules of the road. However, others argue that the longer the GSEs and their legacy securities stay in existence, the longer taxpayer exposure remains. Their proposal would formalize risk-sharing goals and establish a clear schedule and much shorter deadline for the termination of the GSEs.

Beyond unwinding what’s on Fannie and Freddie’s books, what is to happen with the entities themselves? Should they be allowed to continue to exist in this post-reform era? Should they be stripped of their charters? Given that every reform proposal asks for private market involvement in taking on credit risk, a vibrant set of private guarantors will need to take form in the market. Some argue that Fannie and Freddie could fill this role, as they have before, and compete against new entrants to the market.

What is the goal of housing finance policy?

Maybe we should work backwards and ask first what we’d like housing policy to accomplish. Should it increase homeownership rates? Determine an optimal homeownership rate and adjust policy levers accordingly? Increase the supply of rental housing? Ensure that the affordable 30-year fixed rate mortgage lives on? Make housing as plain vanilla as possible or allow for exotic products to exist?

In recent decades, a stated goal of housing policy was to increase homeownership rates. This goal has now been put somewhat to the side, and according to the latest data from the Department of Commerce, homeownership has in fact hit its lowest levels since 1989—63.8%—from a high of nearly 70% in 2005.19 Some believe the final two to three million home mortgages that were issued in the run up to the crisis were the most vulnerable to foreclosure because buyers had the least wherewithal to withstand a downturn. Others disagree and point to other factors far more significant to the housing collapse—like exotic mortgages.

The possible movement of GSE reform legislation serves as an opportunity for policymakers to ask those big questions in housing finance and design a system that reaches those goals.


One of the largest financial regulatory reform packages our country has ever seen, the Dodd-Frank Act, took very few steps to reform Fannie Mae and Freddie Mac. As a result, regulators are acting in a disjointed manner in their efforts to address ongoing issues in the housing finance market and entice private capital to reenter the market. Making matters worse is that the taxpayer is still very much at risk of providing another GSE “capital backstop” should housing markets take a turn for the worse.20

There is no doubt that this recent recession proved that there are risks associated with homeownership. However, studies continue to show that homeownership is the most effective way of helping middle class families build wealth.21 It couldn’t be more apparent that the time to act on GSE reform is soon.

End Notes