Could Fan and Fred go bust?
Perish the thought. Between them the two companies back well over $4 trillion of residential mortgages in the U.S. Underneath this pile of debt the companies have a tiny capital cushion of about $40 billion each. And that counts the $6 billion Freddie recently raised in a preferred stock offering.
Bernard Ducalion did some great work over on Bear in Mind regarding Fannie Mae’s exposure to high risk loans. Though it’s difficult to piece together the current composition of Fannie’s mortgage portfolio–they don’t make it easy, what with varying disclosure in the Ks and Qs they just released–it’s clear Fannie has plenty of exposure to high risk loans.
Try $300 billion of stated-income liar loans (i.e. Alt A), $200 billion of interest only loans, $120 billion of subprime, and $330 billion of >80% LTV loans. They like to play games with the subprime definitions to boot. In the K dated 12/31/06, they have a table that says 5% of their book is loans to borrowers with a credit score below 620. Yet they claim that their “subprime” exposure is a mere 0.3% of the book. I spoke to an IR rep at Fannie, told him that I thought the standard industry definition of a subprime loan is one with a FICO below 620. He said there are other factors that you have to consider and that not all loans with rock-bottom FICO scores should be considered subprime. Really? What WOULD you call a loan made to a borrower with a terrible credit history?
But I digress. The loan categories above aren’t discrete–some high LTV loans qualify as subprime, for instance–so you can’t add them up to get Fannie’s total exposure to high risk loans. But Bernard notes that even if you take a conservative estimate like $400 billion in total exposure, it would take only a 12% fall in value for that chunk of the portfolio in order to wipe out Fannie’s capital entirely.
It’s not quite as simple as that, since Fannie’s book of mortgages isn’t valued on a mark-to-market basis. The decline in value happens more slowly, as credit events like foreclosure actually occur and losses on the mortgages are REALIZED. That’s the way it was explained to me by their rep anyway. Fine. Yet if prices continue to fall as they have been, more and more mortgage vintages will end up upside down and foreclosures will skyrocket, driving huge losses at Fannie.
I replicated Bernard’s Fannie analysis with Freddie. They’re still mired in the accounting scandal that hit both companies so they haven’t filed a Q or K in years. But they do publish a very interesting report on their website that lists their own exposure to high-risk loans. See the larger report dated November 20th, here. Also the “slide presentation” that was released with Q3 earnings.
The two documents suggest that Freddie has $130 billion of Alt A, $150b of interest-only, $70b of loans with FICOs below 620 and $110 billion with LTVs greater than 90%. And these numbers are only for the “guarantee” portfolio; if you include the mortgages that Freddie has purchased, there are another $105b of subprime loans there and $54 billion of Alt A. These last are described as non-agency and the Freddie rep I spoke to said something about how they don’t recognize losses on these particular securities. I was confused by this comment and pressed her: “they’re on your balance sheet. If they have to be marked down as homes go into foreclosure, then the reduction in the value of the asset on the asset-side of the balance sheet must be double-booked as an equal reduction in the company’s equity correct?” Yes, she said. A reduction in equity would be a direct hit to the company’s capital balance.
Again, some loans are in multiple “problem” categories and can’t be double-counted, but you get the idea: clearly Freddie has significant exposure over and above its capital cushion of $40 billion.
I also asked the Freddie rep for clarification on two comments CEO Richard Syron made in the Q3 earnings release. One: they expect $10-$12 billion of credit losses–big, but hardly life-threatening. Two: he anticipates a nationwide drop in house prices of 10%.
My question for IR was whether the loss estimate was based on the housing price estimate. The IR rep confirmed that it was. If house prices decline more than 10%, then their losses will be much larger. Funny how the CEO told investors that the $10-$12 billion loss estimate was pessimistic.
Clearly there wouldn’t be a linear relationship b/t credit losses and house price declines. As prices decline further, a much larger portion of mortgages will be under water. If we get the 30% decline nationwide that some have forecast, it won’t be just the problem loans that cause problems. Plenty of so-called prime loans will be under water and you’ll see “good” credits walking away from mortgages.
If house prices decline 30%, Fannie’s and Freddie’s capital cushions will clearly be wiped out. And taxpayers will have to take on their debts.
Besides being a huge hit for taxpayers. Can you imagine the crisis of confidence that might ensue if these two guys go under? Fully 49% of Fannie’s debt is supplied by foreigners. Fascinating huh? Half of the money Fannie plows into financing the American Dream is provided by foreigners.
It’s possible to see a dollar crisis in that situation as the Fed tries to inflate the economy out of trouble while foreigners punt Fannie/Freddie debt along with other dollar-denominated assets.