In a recent White House memorandum, President Trump said, “It is time for the United States to reform its housing finance system.” He is right about that. He is also right about principal elements of reform, as listed in the memorandum, notably reducing taxpayer risks, expanding the role of the private sector, establishing “appropriate” capital requirements for Fannie Mae and Freddie Mac, providing that the government is properly paid for its credit support of Fannie and Freddie, facilitating competition, addressing the systemic risk of Fannie and Freddie, defining what role they should have in multifamily mortgage finance, and terminating their conservatorships only when the other reforms are put in place.

In all this, the Treasury is instructed to distinguish between what can be done by administrative action and what would require legislation. This seems to set the stage for carrying out the former, even if Congress cannot agree on the latter, which it probably cannot. The memorandum provides that for “each administrative reform,” the Treasury housing reform plan will include a “timeline for implementation.” This timetable instruction represents a pretty clear declaration of intent to proceed.

Needless to say, the general directions can only be implemented after being turned into specifics. While that seems impossible for Congress to agree on, the executive branch can define the specific administrative actions it wishes to take. As the administration comes to decisions about the details, is there an appropriate model to consider for Fannie and Freddie? Is there some way to simplify thinking about the issues they present, which entails swarms of lobbying interests?

I think there is. The model should be too big to fail and systemically important financial institutions. Fannie is bigger than JPMorgan Chase. Freddie is bigger than Citigroup. There is no doubt Fannie and Freddie remain too big to fail. They are an essential point of vulnerability of American residential mortgage finance, the biggest credit market in the world except for Treasury debt. Should they implode, the government will again rush to the rescue of their global and domestic creditors.

Vast intellectual and political efforts have gone into lowering the odds that too big to fail banks will need bailouts or generate systemic crisis. We need to apply the results of that effort to Fannie and Freddie, which now run with virtually no capital. But before the crisis, they already ran at extreme leverage. Indeed, they leveraged up the whole housing finance system, making the system, as well as themselves, much riskier.

What about their capital requirements? Following our model, simply apply the risk based capital standards of systemically important banks to Fannie and Freddie. The same risks need the same capital, no matter who holds them. Fannie or Morgan? Freddie or Citi? The same risk and the same risk based capital. This would result in a required capital for the two on the order of $200 billion, or about $190 billion more than they have.

How much should they pay the government for its credit support? The same as the too big to fail banks pay. For them, this is called a deposit insurance premium. For Fannie and Freddie, you could call it a credit support fee, which would replace the profit sweep in their deal with the Treasury. Deposit insurance fees are assessed on total bank liabilities. Apply the same to Fannie and Freddie credit support fee and at the same level as would be required for a giant bank of equivalent riskiness.

I estimate this would be about 0.18 percent per year, but recommend the administration ask the Federal Deposit Insurance Corporation to run its big bank model on Fannie and Freddie and report on what level of fee results. Note that the Treasury stands behind the Federal Deposit Insurance Corporation, just as it stands behind Fannie and Freddie.

With such a serious amount of capital, Fannie and Freddie would need to charge guarantee fees that would allow greater competition in the private sector. This would be consistent with the law, which requires that their guarantee fees be set at levels that would cover the cost of capital of private regulated financial institutions. If they have the same risk based capital requirement, then that should follow for Fannie and Freddie.

With $200 billion in capital and a credit support fee of 0.18 percent, I estimate that Fannie and Freddie could sustain a return on total capital of 8 percent or so, which is quite satisfactory. The Treasury should exercise its warrants for about 80 percent of their common stock to share in this return until Treasury sells the stock, which will generate a large gain for the taxpayers. It seems to me that virtually all of this might be done by administrative action. Let us hope the administration will proceed apace.

Image credit: Onchira Wongsiri

Featured Publications