Nightmare on GSE Street

Astute observers have long suspected that government-sponsored enterprises (GSEs) like Fannie Mae and Freddie Mac represent a major threat to the U.S. financial system (in fact, I discuss this issue in Chapter 3 of Financial Armageddon).

Of course, when I say "astute," I'm not talking about the highly-paid Wall Street analysts who cover these stocks or the institutional investors who still own millions of shares of either company, because they are basically clueless. I mean people who actually know how to think and who understand such rudimentary concepts as moral hazard and the unintended consequences of government actions.

In "Is Fannie Mae the Next Government Bailout?" Jonathan R. Laing of Barron's lays out out the bear case for this particular disaster-in-the-making, though his analysis does fall short in one respect. By the time any bailout does come about (assuming it does), the financial system will have been shredded like Freddy Kreuger's unwitting victims in the Nightmare on Elm Street film series.

It's perhaps the cruelest of ironies that in the U.S. housing market's greatest hour of need, the major entity created during the Depression to bring liquidity to housing, Fannie Mae, may itself soon be in need of bailout.

Fannie, of course, occupies a curious middle ground between the public and private sector as a result of its privatization in 1968 as a Government Sponsored Enterprise, or GSE. While owned by its shareholders, Fannie is regulated by a government agency and is able to borrow money cheaply, thanks to an implicit guarantee by Uncle Sam. It uses those funds to buy and securitize home loans -- lots of them. At year end, the company owned in its portfolio or had packaged and guaranteed some $2.8 trillion of mortgages or 23% of all U.S. residential mortgage debt outstanding.

Of late, however, Fannie's prospects have darkened notably. The company (ticker: FNM) lost $2.6 billion last year as a surge of red ink in the final two quarters more than wiped out a nicely profitable first half. And by late last week, credit-market jitters had penetrated the once-unassailable hushed precincts of the market in Fannie debt.

In the wake of margin calls on collateral at the investment concern Carlyle Capital, yields on guaranteed mortgage securities issued by Fannie and its GSE sibling Freddie Mac (FRE) rose to their highest level over U.S. Treasuries in 22 years. Likewise credit default swaps, measuring market concerns over the safety of Fannie corporate debt, have ballooned out to 2% of the insured amount from 0.5% just four months ago.

Company executives attribute such concerns to what Fannie CEO Daniel Mudd last month called "the toughest housing and mortgage market in a generation." He also said that much of 2007's loss came from reducing to market levels the value of derivatives that Fannie uses to hedge its interest-rate risk. And those accounting moves should reverse and fatten earnings in the fullness of time once interest rates stop dropping.

But, if the truth be known, a considerable portion of Fannie's losses also came from speculative forays into higher-yielding but riskier mortgage products like subprime, Alt-A (a category between subprime and prime in credit quality) and dicey mortgages requiring monthly payments of interest only or less. For example, Fannie's $314 billion of Alt-A -- often called liar loans because borrowers provide little documentation -- accounted for 31.4% of the company's credit losses while making up just 11.9% of its $2.5 trillion single-family-home credit book. Fannie was clearly looking for love -- and market share -- in some of the wrong places.

Likewise, Barron's has found other areas that may bode ill for Fannie's prospects. Its balance sheet is larded with soft assets and understated liabilities that would leave the company ill-equipped to weather a serious financial crisis. And spiraling mortgage defaults and falling home prices could bring a tsunami of credit losses over the next two years that will severely test Fannie's solvency.

Should Fannie or the similarly hobbled Freddie Mac buckle, the government would no doubt bail them out and honor their debt and mortgage guarantee obligations. Fannie common and preferred shareholders would likely suffer grievously in such a scenario.

Fannie, for its part, insists it's more than adequately capitalized to withstand any future stress. The company also contends that as a result of tightening its standards and making fewer risky loans, the quality of its book of business will improve mightily.

But some financial leaders aren't so sure. At a conference several weeks back, William Poole, president of the St. Louis Federal Reserve Bank, said that the GSEs (clearly a reference to Fannie and Freddie) appeared to be insufficiently capitalized to handle the kind of losses suffered by U.S. major banks in the past six months. "I do not have any information on the GSEs that the market does not have," he said. "Nevertheless, in assessing the risk of further credit disruptions this year, I would put the GSEs at the top of my list of sources of potentially serious trouble."

And, in commenting on the government's "too big to fail doctrine" for financial institutions, he said: "First, firms in trouble ought not to be bailed out, unless the bailout takes a form that imposes heavy costs on managers and shareholders."

Poole has long been skeptical -- correctly it turns out -- of Fannie and Freddie's ability to serve both God (their social mission of promoting liquidity and affordability) and Mammon (the shareholder and lush management compensation). At Fannie, a generation of Democratic Party insiders, such as James Johnson, Jamie Gorelik and Franklin Raines, made substantial fortunes in Fannie's executive suite. As Fannie Mae's top regulator, James Lockhart, pointed out in recent congressional testimony, the absence of debt-market discipline (the government guarantee makes Fannie and Freddie all but impervious to credit downgrades) makes pell-mell growth irresistible to shareholders and managers. Have a hunch, bet a bunch.

A major scandal erupted at Fannie earlier in the millennium when the company was found to be cooking its books to hide a multibillion-dollar loss it had incurred when massive interest-rate bets went awry. Freddie got nailed at the same time for setting hedging profits aside in a cookie jar to boost results in subsequent years. Yet, the recent lending bets made by Fannie are likely to prove far more damaging.

On the surface, Fannie's balance sheet looks fine. At year end, the company reported regulatory net worth of $45.4 billion, some $3.9 billion higher than the expanded minimum capital of $41.5 billion required by federal regulators. But with its extreme leverage -- assets stand at 20 times net worth -- Fannie has little room for error. And there appear to be significant problems with the way Fannie has valued both its assets and liabilities.

For example, some $13 billion of its $45.4 billion in net worth consists of deferred tax assets that have value only if Fannie can earn enough money in the near future (say $36 billion) to employ them. That hardly seems likely. During the housing boom of 2002 to 2006, this tax asset only climbed -- from zero to $8 billion as Fannie reported $23 billion in income from 2003 to 2006.

Last year's $2.6 billion loss compounds the problem, pushing the tax asset to $13 billion. At a minimum, accountants may require the company to sharply write down the value of this asset, thus slashing net worth. Bank regulators, for example, limit the amount of deferred tax assets for regulatory purposes to the lesser of the amount expected to be used within one year or 10% of regulatory capital. So if Fannie were a bank, this entire asset would be wiped out. Fannie maintains the value of the asset will be realized over time.

Another soft asset is Fannie's $8.1 billion of Lower Income Housing Tax Credit partnerships. The partnerships' only value, other than helping fulfill Fannie's housing affordability requirements, are the rich tax credits they generate from their intended operating losses. The problem is that Fannie hasn't made enough money to employ these tax credits. Thus the asset is apt to dwindle away to zero without providing Fannie any benefit. Fannie makes no predictions on the future values.

The story is much the same for the liability side of Fannie's balance sheet. There's an item called guaranty obligation, which represents the company's best estimate on what it will have to pay out to make good on any mortgage defaults in its $2.4 trillion guaranty book. On its regular balance sheet, Fannie carries the item at $15.4 billion, but on its "fair value" balance sheet, which attempts to mark every asset and liability to current market value, the guaranty obligations are pegged at $20.6 billion. The problem was, as Morgan Stanley analyst Kenneth Posner discovered, Freddie went through the exact same drill with its guaranty obligations' fair value and chose to mark them much more aggressively. It valued them at 1.5% of its guaranteed book, double the 0.74% of total book that Fannie saw fit to use, even though Freddie's delinquency rate is lower than its rival's.

Had Fannie taken a similar hit, its fair-value net worth would've shrunk by some $20 billion to a paltry $16 billion, compared with its juiced-up regulatory capital of $45.4 billion. Fannie stands by its estimate and says it doesn't know how Freddie arrived at its own.

Finally, Fannie seemed to have been inordinately easy on itself when, in the fourth quarter, it wrote down its $74 billion holdings of privately packaged, non-agency subprime and Alt-A mortgage securities by a mere 6%, or $4.6 billion. In addition, Fannie declared that only $1.4 billion of the write-down constituted a permanent impairment, something that penalized both Fannie's profits and net worth. The remainder of the write-down was deemed a temporary mark-to-market loss that had no such negative impact.

Had Fannie charged off the remaining $3.2 billion that would have torched most of the $3.9 billion in excess regulatory capital that it held at the end of the fourth quarter. Nearly all the major banks, from Merrill Lynch to UBS, have taken much larger percentage write-downs on their holdings of similar mortgage paper, and ran virtually all the losses through their income statements.

In any event, continued deterioration since year end in indexes like the ABX triple-A index indicate that Fannie, based on the different vintages it owns, should conservatively take another $14 billion charge, according to Barron's estimates. Fannie Mae says that since it's a long-term investor, it should incur no permanent decrease in asset value beyond what it has recognized.

The very survival of Fannie as a going concern hinges on the size and speed of the credit losses it faces in the years ahead. Merrill Lynch's Kenneth Bruce sees Fannie suffering losses on its current book of around $32 billion over the next decade. Yet, he still expects the company to manage recovery earnings per share of between $2.50 to $4 between 2009 and 2011.

His forecast, however, is based on spirited 8% average annual growth in Fannie's credit book over the decade. Although Fannie has just been cleared to deal in mortgages of up to $700,000, from $420,000 now, 8% growth could be hard to come by if the company's capital remains stretched.

In our view, the rapid decline in home prices and soaring level of foreclosures might cause the wave of credit losses to hit far sooner and with greater ferocity than many imagine, potentially submerging the income Fannie is expecting to harvest from volume growth and higher lending fees.

Click here to read the entire article (subscription required).

Comments

[Michael Fernwood]

Maybe you can help. No one seems to have asked this question, let alone answered it.

In the quote above, Fannie's $314BB in Alt-A which accounted for 31.$% of their losses but only 11.9% of their credit book. Is that $314BB made up of private issue Alt-A MBS in their investement portfolio, or is it in their "originated" portfolio (either cash sale whole loans or MBS)?

Fannie and Freddie's Automated Underwriting Systems approved a lot of liar loans. (An interesting aside about liar loans that everyone seems to miss, is that you were able to lie about your assets, also). As DU and LP got more "freaky" with their approvals, lender Alt-A guidelines expanded. About the only difference is that an Alt-A book probably skews towards a higher average loan amount and lener Alt-A charged a rate premium for the risk. Fannie and Freddie didn't.

[Michael Panzner]

I don't know the answer to your question. You might want to send an email to the reporter who wrote the story, Jonathan R. Laing, at jon.laing@barrons.com.