Mark Zandi: A More Promising Road to Reform
April 7, 2016
In today’s housing finance system, two behemoth institutions, Fannie Mae and Freddie Mac, control most of the core infrastructure of the secondary market and take on most of its credit risk. While in many ways this system has served the nation well by providing a broad range of borrowers access to credit and a level playing field for lenders of all sizes, our reliance on this duopoly created perverse incentives that ultimately led to too much risk taking, forcing taxpayers to shoulder the resulting cost.
Repeated attempts to reform this system have failed, none able to strike an adequate balance between overhauling what hasn’t worked and preserving what has. So we remain in a state of limbo that no one believes is sustainable, looking anxiously for a way out.
A group of us has thus put forward a reform proposal that addresses the system’s structural problems and keeps what works. Our proposal moves the core infrastructure of the secondary market from Fannie and Freddie into a government corporation—the National Mortgage Reinsurance Corporation (NMRC)—and transfers all the noncatastrophic credit risk into the private market, where there is competition and innovation without risk to the taxpayer. In our proposal, no private institutions dominate the housing finance system by controlling its infrastructure or taking on the lion’s share of its credit risk. The result is a system that will function like the current one for both borrowers and lenders, but without the underlying risk posed by a dependence on “too-big-to-fail” institutions.
Those who find this idea appealing should read "A More Promising Road to GSE Reform," where the proposal is described in detail. What follows here is a summary of some of our proposal’s key features and comparisons with other reform proposals.
Our proposal ends the housing finance system’s reliance on too-big-to-fail financial institutions by putting the core infrastructure on which mortgage market participants depend into a government corporation and transferring the credit risk across a diverse range of private sector actors, including the capital markets, reinsurers, private mortgage insurers, lenders, and other private entities. No private institution would be indispensable to a healthy, functioning secondary market; any institution could fail, and the system would not be affected.
Other reform proposals rely on too-big-to-fail private institutions in one form or another. Recapitalizing Fannie and Freddie and privatizing them would re-create this problem in the form we had before the crisis. Systems that replace Fannie and Freddie with a shareholder-owned utility or a mutually owned institution are a vast improvement, but those systems leave this critical flaw unaddressed. And while proposals that replace Fannie and Freddie with multiple private guarantors solve many of the challenges in today’s system, these proposals are unlikely to establish enough guarantors that a regulator would ever let any of them fail. Each alternative proposal thus leaves in place some version of the incentives that led to the failure of the current system.
Maintaining access for borrowers and lenders
By requiring the NMRC system to meet affordability and duty-to-serve goals defined by the Federal Housing Finance Agency (FHFA), we assure the same broad access for underserved communities that is required of Fannie and Freddie today. To achieve these goals, the NMRC would maintain the same underwriting standards and practices as Fannie and Freddie, and would price its guarantee fees in a manner that subsidizes lower-wealth borrowers who are creditworthy but may not be able to afford a mortgage loan otherwise.1 In addition, an explicit affordability fee would fund initiatives to support access and affordability for homeownership and rental housing.
Community banks and small lenders would have the same access to the system they have today with Fannie and Freddie, namely through the NMRC’s cash window. Moreover, the NMRC’s mandate to provide broad, competitive access to the secondary market would ensure that lenders of all sizes have the same access to the market’s infrastructure and benefit equally from the NMRC’s credit-risk transfers.
The NRMC would maintain the benefits of today’s system at almost no risk to taxpayers. The private capital in the NMRC system would be on the hook for the first 3.5 percent of loss, consistent with the current implied capitalization of Fannie and Freddie and consistent with the losses suffered by the government-sponsored enterprises (GSEs) during the financial crisis. In case there is an even more severe cataclysm in the future, a mortgage insurance fund would cover the next 2.5 percent of loss. And in case this doesn’t cover it, the FHFA, the regulator in the NMRC system, would have the authority to claw back any losses taxpayers suffered by charging a higher insurance fee after the crisis passes and the economy normalizes. No other reform proposal offers more protection to taxpayers.
Providing low and less cyclical mortgage rates
The NMRC system can provide this formidable taxpayer protection without driving up mortgage rates. There are some new costs in the NMRC system, including a fee for the government’s explicit catastrophic reinsurance and an affordability fee for funding activities to support access for underserved communities. But these costs would be offset by lower yields on NMRC mortgage-backed securities (MBS). Unlike Fannie and Freddie’s MBSs, the NMRC’s MBS would be explicitly backed by the full faith and credit of the US government and would thus trade more like Ginnie Mae’s explicitly guaranteed MBS, which trade at a much lower yield than Fannie and Freddie’s MBS.
Because there are no too-big-to-fail institutions in the NMRC system, the amount of capital required to support the system would be lower than in other systems. Less required capital means lower costs and thus lower mortgage rates. How much lower depends on what additional capital regulators would require the too-big-to fail institutions in the other systems to hold.
Moreover, despite the reliance on private capital in the NMRC system, mortgage rates may actually be less cyclical than in the current system. Guarantee fees would be more cyclical, as Fannie and Freddie’s current guarantee fees rarely change in response to market conditions, but NMRC’s guarantee fees would vary depending on the cost of private capital, which in turn would fluctuate with the perceived risk in the market. However, changes through the business cycle in other components of mortgage rates, such as the risk-free interest rate and lenders’ margins, would likely offset the changes in guarantee fees. In good times (bad times), the cost of capital and guarantee fees would likely be lower (higher) in the NMRC system than in the current system, but the risk-free rate and lenders’ margins would be higher (lower).
Even the cyclicality of the guarantee fees in the NMRC system would be mitigated by the use of multiple sources of capital. A critical feature of our system is that it gradually migrates toward a mix of risk-transfer structures shown to be resilient through the business cycle. We envisage, for instance, insurance companies and other entity-based sources of capital, who assume credit risk through good times and bad, to fully participate in the risk transfers. For how we think about this and other key objectives of credit-risk transfers, see "Delivering on the Promise of Risk Sharing."
The NMRC would also have the authority to ease the impact on rates and liquidity during a financial crisis. In a time of acute stress, private investors would be unwilling to provide capital or require such a high return that it would cause guarantee fees and mortgage rates to spike, exacerbating the financial turmoil. To ensure that this does not happen, the NMRC could scale back its risk transfers when private capital’s required return rises above a predefined crisis threshold.
Promoting flexibility and competition
As a government corporation, the NMRC would have considerable flexibility. It would not face the same constraints in rulemaking or employee compensation as a government agency, for instance, or depend on Congress for funding. This would allow the NMRC to function with more of the flexibility of a private entity, which would be critical in managing an infrastructure as complex and fluid as our housing finance system.
It is unlikely that systems based on a privately owned mutual or utility would be able to provide significantly more flexibility. The institution would be a heavily regulated monopoly whose range of business activities, rate of return, and market share would be closely prescribed by policymakers, removing many of the incentives and constraining much of the flexibility that drive the typical private company to be more innovative and efficient.
By putting the market’s infrastructure into a government corporation, lenders of all sizes would have the same access. The system would not be beholden to larger institutions that gain an advantage by controlling access to the secondary market, as has previously occurred and would be repeated in many alternative proposals. There would thus be no barriers to entry into the primary market, leveling the playing field and increasing competition.
The NMRC system would also promote competition in providing private capital to the system by utilizing a wide range of sources, such as the capital markets, insurance companies, and other private entities. Moreover, the plethora of credit-risk transfer structures offered by Fannie and Freddie would eventually be rationalized into a mix of structures chosen to ensure a level playing field for all lenders, along with broad access and stable liquidity through the business cycle.
Easing the transition
Finally, under our proposal, the transition from the current system to a much-improved one would occur with little disruption, uncertainty, or risk, and would build upon steps already under way. Fannie and Freddie would continue to build the common securitization platform, and the current effort to synchronize some of their processes would gradually be extended to all of them, from purchasing mortgages to securitizing them and overseeing their servicing. And Fannie and Freddie’s current risk-transfer efforts would gradually be expanded from an effort to transfer some noncatastrophic risk to the private market to one in which they transfer all of it.
It is critical to move in this incremental fashion, as the structural reform called for here requires changes to a remarkably complex and important system. Rather than mandate today which capital structures should cover losses ahead of the taxpayer, for instance, it is better to mandate what they must achieve and allow the NMRC and the market enough flexibility to develop the mix of structures that is shown over time to best achieve it. With this in mind, Fannie and Freddie—and ultimately the NMRC—will gradually shift their risk-transfer efforts to the most effective mix of structures. Once Fannie and Freddie are issuing a single security from a single platform, operating under a single set of processes, and syndicating all of their noncatastrophic credit risk, their operational assets will be put into the newly formed NMRC.
In other proposals that offer the significant structural reforms required to improve the system, the transition process is inevitably much more daunting; they require creating privately held institutions that control the market’s infrastructure and manage credit risk. Whether it’s creating new too-big-to-fail guarantors, a mutual with thousands of members, or a single, privately owned utility, the new system’s ability to bear the weight of the mortgage market in a way that achieves all of our objectives depends on whether policymakers appropriately gauge the market for investing in a complex, highly regulated system.
Like all reform proposals, ours ultimately requires legislation for the system to receive the full catastrophic backstop of the US government. But since our system sheds the problems and keeps the benefits of the current system in a way that minimizes the risk and uncertainty of transition, it should appeal to lawmakers on both sides of the aisle.
The future is now
The NMRC system is vastly preferable to the current system and offers important advantages over the alternatives. You may agree with us or, instead, prefer another path toward reform offered in this Urban Institute series, but it is critical that lawmakers re-engage on this issue. The only question is whether we have the discussion now, in a moment of relative calm, or later, when a crisis forces our hand.
- One reviewer of our proposal appears to have misunderstood our limitation on the loans that the NMRC will buy, pool, and securitize to those that meet the product features of a “qualified mortgage,” as defined by the Consumer Finance Protection Bureau. With this limitation, we would only preclude loans with the product features prohibited in the rule, such as interest-only loans, negatively amortizing loans, and loans with balloon payments. We would not preclude loans on the basis of debt-to-income ratio. For more on this topic, see “What is a Qualified Mortgage,” by the Consumer Financial Protection Bureau, updated February 8, 2016.