It’s official. After the administration’s protestations to the contrary last week – read my lips, no new bailouts – this morning Treasury announced a plan to inject “billions of dollars in loans and investments” to shore up Fannie and Freddie.
Let’s remember how Freddie and Fannie work, and financially what is at stake for taxpayers.
Freddie’s and Fannie’s main business is to buy mortgages from financial institutions that lend money to homeowners. They are shareholder-owned public companies, but as “government sponsored entities,” they have special charters and legal preferences that reinforce investors’ perception that their debt securities bear an implicit federal guarantee.
Freddie and Fannie hold about $1.6 trillion of these mortgages on their own balance sheet. They insure an additional $3.6 trillion of mortgages, which they resell them as mortgage-backed securities. There are two main risks that result from these activities.
First, Fannie and Freddie issue debt—- known generically as “agency debt”—- in order to raise the cash to pay for the mortgages held on balance sheet. The debt they issue is typically fixed-rate debt, while the mortgages they hold have an uncertain maturity. They engage in derivative transactions to partially hedge the resulting cash flow mismatch between the mortgages they own and the bonds they issue. To the extent these hedges fall short they are exposed to losses from interest rate and prepayment risk.
Second, Fannie and Freddie bear the risk of default losses on these mortgages. For the most part, Freddie and Fannie have bought only conforming mortgages, for which homeowners put 20% down or obtain insurance, and for which they verify their income and employment.
With the down payment cushion, steadily appreciating home prices, and a nationally diversified portfolio, for many years most observers played down the importance of credit risk for Fannie and Freddie. Recently, however, with house prices falling rapidly, the risk that large numbers of homeowners will default on their mortgages has become a reality.
Capital available to absorb losses is razor thin. Fannie and Freddie bear credit risk on over $5 trillion of mortgages. Their combined regulatory capital is about $80 billion, and their market capital is less than a third of that.
How much could this cost taxpayers? First, look at real estate prices. The Case-Shiller housing price index (the last available is for April) has lost 15% of its value over the past year, and 20% since the peak of real estate prices. In many cities the loss is close to 30%. Since this is an average, some houses will have lost less, some more. Those losing the most are the likeliest to default. Let’s consider some scenarios.
Suppose that 25% of the mortgages insured by Fannie and Freddie experience an average loss of 10% (with a 20% down payment, this would mean that the underlying houses on average lost 30%. The group that defaults would have a significant loss; otherwise the homeowner would try not to default). This is a 10% loss on $1.25 trillion, or $125 billion. If 40% of mortgages default and the loss on those mortgages is 20%, the loss for Fannie and Freddie is $400 billion.
Given these numbers, a loss in the tens of billions seems likely. It is not hard to imagine a $100 billion loss, while a $400 billion loss seems much less likely. Reports are that the Treasury will ask for the right to extend a $300 billion line of credit to Fannie and Freddie. Given these back-of-the-envelope calculations, this seems a reasonable amount. The bottom line is a painful reminder that “implicit guarantees” can be extremely costly, but these losses will not bankrupt the government.
An important question that is harder to answer is whether the build-up of easy credit that created tinder for the current credit meltdown was driven by the extraordinary ability of Fannie and Freddie to raise mortgage capital at low cost and with minimal oversight. This is a topic for another day.