Fannie Mae & Freddie Mac: Money Well Spent?

On Sunday, Washington policymakers announced details of a plan to shore up Fannie Mae and Freddie Mac, the struggling government-sponsored enterprises that are huge players in the mortgage market. The proposals—a plan for both the Treasury Department and the Federal Reserve to extend credit to the listing giants, if needed, and a proposal for Congress to give Treasury the authority to buy shares in the companies, if needed—were immediately characterized as a bailout.

But the actions don't amount to a bailout—yet. The announcements simply constituted the final explicit declaration of what investors, politicians, and policymakers around the world have long taken as an implicit assumption: that the U.S. government would back the debt of Fannie Mae and Freddie Mac. Questions have been raised about the potential cost to taxpayers if the government ultimately does have to help the two companies make good on the debt they've sold to investors. But as it turns out, under almost any circumstances, the bailout will be a bargain for American taxpayers, because any cost of it will be overwhelmingly offset by the tangible and quantifiable economic benefits that taxpayers have collectively received over the years from the market's expectations that such a bailout would materialize if needed.

Fannie Mae and Freddie Mac occupy a strange netherworld between the U.S. government (whose debt can never fail) and the corporate world (where debt occasionally fails). Fannie and Freddie borrow money in the public markets at rates somewhere between what the government pays and what a good corporate borrower would. Over the years, economists have attempted to quantify the government-sponsored enterprise's advantage—and how much of that advantage it passes on to borrowers in the form of lower costs. On its Web site, Freddie Mac quotes the Office of Management Budget thus: "[M]ortgage rates are 25 – 50 basis points lower because Fannie Mae and Freddie Mac exist in the form and size they do." (Twenty-five to 50 basis points is one-quarter to one-half a point, in layman's terms.) Freddie goes on to say that "because the secondary mortgage market saves homebuyers up to one half percent on their mortgage, borrowers nationwide save an average of nearly $23.5 billion annually." That may be overstating the case. Economist Lawrence J. White of New York University's Stern School of Business says the consensus among economists holds that the implied guarantee allows Fannie and Freddie to borrow at rates between 35 and 40 basis points lower than rates available to analogous companies that don't have an implicit government backing. If Bank X pays 5.35 percent to borrow, Fannie Mae pays only 5. The government-sponsored enterprises pass on most—but not all—of those savings to consumers in the form of lower interest rates on mortgages. Again, White says the consensus is about 25 basis points. (Studies by Wayne Passmore of the Federal Reserve, seen here and here, put it lower. This article, which reviews several other studies, put the savings at only seven basis points.)

But 25 basis points—one-quarter of one percentage point—can add up to significant savings on large amounts of money. At the end of 2007, Fannie and Freddie held or guaranteed mortgages worth about $5.2 trillion. (The annual report of OFHEO, the regulator of Fannie and Freddie, has great historical data on the two companies. Information on their balance sheets can be seen in Table 22.) This means that borrowers saved about $13 billion in 2007 on interest costs, thanks to the GSEs. The numbers were lower in previous years. Assuming a 25-basis-point savings, the savings were $7.5 billion in 2001 and $3.25 billion in 1994. Calculating the cumulative present-day value of the historic implied guarantee—after all, $3.25 billion in 1994 is worth about $4.75 billion in today's dollars—would be a complicated task. But after laying out all sorts of caveats, Lawrence White believes it could add up to more than $100 billion in current dollars.

So, merely by signaling to the markets that it might back the GSEs' debt, the government has, over the past few decades, helped tens of millions of homeowners save some serious money. Until this week, the cost of this benefit has been effectively nothing—save for some foregone taxes and the cost of regulating the companies. But if Fannie and Freddie exhaust the patience of the private sector—the shareholders—and can't raise capital to make up for losses on the mortgage portfolio—they would have to turn to the government. How much would they need?

This is a great unknown. But it's hard to imagine it would approach a figure close to $100 billion. Fannie Mae and Freddie Mac didn't make subprime loans, although they do have some exposure to subprime debt through assets they purchased. Rather, they make loans to people who make down payments and who buy houses under a certain price (the maximum loan last year was $417,000). As a result, the companies avoided funding lots of mortgages in expensive, bubbly markets. In the fourth quarter of 2007, the delinquency rates for mortgages on single-family homes were 0.65 percent for Freddie Mac and 0.98 percent for Fannie Mae.

Let's assume for the moment that 5 percent of the $5.2 trillion in mortgages that Fannie Mae and Freddie Mac hold or insure goes bad, which would represent a massive (and unlikely) uptick from current numbers. Lawrence White says that because of the company's underwriting standards, the losses on those loans would be only about 30 percent. Run the numbers, and the potential losses—i.e., the amount of federal funds needed to make bond investors whole assuming the GSEs can't raise any more outside capital—would be about $78 billion. If 2 percent of the mortgages the GSEs hold or insure goes bad—a much more reasonable guess—the government would have to come up with about $31 billion. Those are big hypothetical costs for taxpayers. But they would still be smaller than the actual benefits taxpayers—or at least the large majority of taxpayers who are homeowners—have already received.