Just How Risky Are Fannie and Freddie?

Not very, says a new report. But it probably won't sway the mortgage giants' critics

For years large commercial banks have urged Congress to rein in the galloping growth of mortgage-finance giant Fannie Mae (FNM ) and its smaller cousin, Freddie Mac (FRE ). So far the two government-chartered, shareholder-owned companies have fought off efforts to restrain them. But the "too big, too risky" refrain grew louder in the past year as Federal Reserve Chairman Alan Greenspan and Treasury Secretary John W. Snow added their voices to the chorus. They advised Congress to put Fannie and Freddie under Treasury's oversight and let regulators set capital standards -- a backdoor way to limit the housing giants' growth.

Not everyone is as worried as Greenspan and Snow -- particularly R. Glenn Hubbard, President George W. Bush's former chief economist. In a 33-page study that Fannie Mae commissioned and made available to BusinessWeek before publication, Hubbard, who now heads the Columbia Business School, argues that Fannie is no riskier than the 10 largest commercial banks or their holding companies. Hubbard also concludes that Fannie's capital reserve -- the cushion that financial institutions need in case of emergency -- is adequate.

Why does this matter? Fannie Chairman and CEO Franklin D. Raines will use the report to rebut critics who charge that Fannie's low borrowing costs represent an implicit federal subsidy created by debtholders' confidence that the government will rescue them in a financial crisis. Hubbard's work will bolster Raines' argument that Fannie has a cost advantage because it's low-risk. In the study's foreword, Raines writes that Hubbard "confirms the low-risk nature of Fannie Mae's business but also...indicates a fundamental flaw in studies that purport to quantify an 'implicit subsidy' to the company." Hubbard says: "That's Frank Raines's view."

However it's spun, the study is the first real measure of Fannie's risk to its bondholders and the government. It may well intensify the debate over Fannie's future. Treasury and Fed spokesmen refused to comment on the paper. But other experts say Hubbard's quantitative approach fails to consider the everyday stresses and strains Fannie faces.

"This is a theoretical, academic exercise that doesn't capture Fannie's real-world risks," says Michael T. DeStefano, managing director for financial institutions at Standard & Poor's (MHP ).

In some ways, Hubbard's approach is like the mathematical arguments used by Long Term Capital Management to claim that its risks were properly hedged, right up to its 1998 collapse. But while Hubbard concedes that his analysis is strictly quantitative, his model considers perfect storm scenarios, including a one-year jump in rates to 11%, a 5% drop in home prices nationwide, and up to a 25% fall in regional housing. And unlike LTCM's illiquid assets, Fannie holds easily saleable home mortgages.

Still, Hubbard does not consider scenarios that can't be quantified because they have never happened, such as the systemic risk of a rapid interest-rate increase that causes housing prices to decline, touching off a series of financial-institution failures. In response, Hubbard says that risk studies aren't usually based on cataclysms for which a bank would have to hold infinite amounts of capital. "You're not typically looking for events that aren't in the record," he says. Hubbard's analysis rests on econometric models that crunch factors including the quality of Fannie's assets, its exposure to sudden changes in interest rates, the volatility of its earnings, and the probability of default. He concludes that, if Fannie failed, the government and bondholders could expect to lose $90 billion. That's just 8.9% of Fannie's $1 trillion in assets, compared with the 22.3% of assets that would be lost if the typical commercial bank failed. Neither Hubbard (who also writes a column for BusinessWeek) nor Fannie would reveal how much he was paid for the analysis.

But Hubbard's study isn't likely to silence Fannie's critics. One reason is that Hubbard views Fannie's high level of concentration -- 90% of its balance sheet is mortgage-based -- as a plus, since losses on mortgages have been historically low. But S&P's DeStefano worries about the lack of diversification. That means Fannie has no other business to rely on if housing deteriorates.

DeStefano, one of few experts who has read the study, also argues that the bond market accepts a lower yield from Fannie because "everybody looks at Fannie as though it's part of the government," and no study can erase that perceived link. If Fannie were to lose its implied government backing, borrowing costs would probably go up -- if only because S&P would reconsider its triple A credit rating, he says. S&P, like BusinessWeek, is a unit of The McGraw-Hill Companies (MHP ).


Another economist who read the study and is familiar with Fannie also takes issue with Raines's conclusion that Fannie's lower borrowing costs are due to its low-risk profile. Fannie's big advantage, he says, is its ability to act like a mini-Treasury: Unlike commercial rivals, it can issue debt securities quickly and at will, without having to register the debt with the Securities & Exchange Commission. Investors accept a lower rate on Fannie's debt because they know they'll always find a buyer. Fannie owes that liquidity to its quasi-governmental status, not to lower default risk, says this economist, who asked not to be identified.

One more problem: Hubbard doesn't mention the legal limbo Fannie's debtholders would be in if Fannie were to fail. Fannie's regulator, the Office of Federal Housing Enterprise Oversight, has no authority to act as a receiver and determine the order in which claimants would be paid. With no receiver, Fannie's debtholders might trigger panic selling (GE ).

What worries Greenspan and Snow is the risk of financial contagion based on unforeseeable events. They're also alarmed by Fannie's high debt: $962 billion at the end of 2003. They and other critics aren't likely to be swayed when Raines asserts that Fannie's capital reserves are adequate. Hubbard may add a scholarly layer to the discussion, but he probably won't change many minds.

By Paula Dwyer in Washington

    Before it's here, it's on the Bloomberg Terminal.