“… both the Fed and the government are afraid that they might have to make good on a guarantee that doesn’t exist.”

On April 13, 2005, Toronto Globe and Mail technology reporter Mathew Ingram used those words to close his story “Don’t be fooled by the cute names: U.S. Treasury Secretary John Snow is worried by the size of GSE retained portfolios”. Fifteen months later the US has a new Fed chairman and a new Treasury Secretary, but the words still ring true. Only now the risk is immediate.

America’s trademark 30 year fixed rate repayable mortgage is the envy of the world, but it comes at a price. If rates go up, homeowners continue paying their lower fixed rate and the bank may need to refinance at the new higher rates. If rates go down, the homeowner can pay off the high mortgage and refinance and the bank again loses. A lot of things can happen over 30 years, so the bank must prepare for the possibility that the homeowner might get into trouble at some point and default. No local bank can manage this 30 year long package of risks for a large number of mortgages in their area, so the government created quasi-private companies to help.

“Fannie Mae — whose formal name is the Federal National Mortgage Association — and Freddie Mac (whose real name is the Federal Home Loan Mortgage Corp.) were created to help Americans buy new homes, first in the years following the depression and then in the Baby Boom years. They accomplished this goal by buying mortgages from thousands of banks throughout the United States and then packaged them and sold them to investors, in a process called ’securitization.’” (Ingram 13Apr05)

Fannie Mae and Freddie Mac are the two largest institutions that perform mortgage securitization for US banks, and they are huge. Also performing similar functions are a dozen Federal Home Loan Banks (FHLBs) and additional institutions that all come under the category of US housing government-sponsored enterprises (GSEs). When the GSEs buy mortgages from local banks, the local banks are relieved of the risks from these mortgages and can issue more. So the whole point of the exercise is that the GSEs assume all the risk for most the homeowner mortgages in the US financial system. Securitization creates bonds called Mortgage Backed Securities (MBS), and these bonds, especially the senior debt that comes out of Freddie and Fannie, is given the generic name of US agency debt. GSEs are really, really serious about risk management because that is the essence of their business, but things could still hypothetically go wrong. One thing that could go wrong is millions of Americans going upside down in their mortgages over the next couple of years due to house prices collapsing in selected markets and ARM resets. Recent news stories and analyses suggested this scenario is not unthinkable. This sounds like a scary situation for the investors, but agency debt bondholders have always thought they had an additional ace up their sleeves.

“…the market seems to be betting that Fannie and Freddie are so gigantic, and so crucial to the overall framework of the U.S. financial markets — since most banks hold a lot of their debt — that there’s no way the government would allow them to fail. (Ingram 13Apr05)”

This is called the implicit guarantee, and is that same “guarantee that doesn’t exist” Ingram mentions in the quote at the top of this post.

Freddie and Fannie between them have issued somewhere around one point eight trillion dollars in agency debt, and Congress certainly does not want to take responsibility for it should the big GSEs get in trouble over a future collapse of the housing bubble. In fact, Freddies’ 1972 charter may have been explicitly designed to privatize MBS debt in the aftermath of Nixon closing the gold window in 1971, making the US dollar an inherently delicate “fiat currency”. As a matter of fact, the actual MBS bonds all come stamped with big notices to the effect that The US Government DOES NOT BACK THIS BOND. And more and more officials and analysts are trying to get the message across that they aren’t kidding. A recent example is this 13Jul06 Christian Science Monitor op-ed “No safety net for Fannie and Freddie: In the wake of scandals, Congress should follow the Bush administration to take bolder action”, By David Reiss. The bottom line is a practical consideration: this is a very bad time for Congress to admit that their debt obligations are hundreds of billions or trillions of dollars greater than previously admitted. But this would be the cost of a successful invocation of the implicit guarantee.

That outlines what might happen should the big GSEs collapse. So how are they doing in the risk management business anyway? On the positive side, they have employed some of the world’s best economists and computer models to perform this core function. Some years ago a Nobel Prize winning economist wrote a report concluding that there was less than a 1 in 100,000 chance that Fannie would run into trouble, given the safeguards in place. To balance this, both Fannie and Freddie have recently had significant problems with the SEC and their regulator, the Office of Federal Housing Enterprise Oversight (OFHEO). They have had to make significant revisions to their earnings, and have not yet published good audited financials covering the last several years (Fannie is not expected to publish for at least a couple of years yet). That makes it a hard to analyze the present state of their risk management configuration.

In future posts, I will examine GSE risk management in more detail and point to places where critics of Freddie, Fannie, and their friends have provided analyses and suggestions for improvements. In the meantime, you might want to check out arguably the most important of these efforts, the Housing GSE Project at the neoconservative American Enterprise Institute (AEI).

—–

John McLeod grew up in 02420, half a mile south of Route 128 through the 1950s and 1960s. For the last quarter century he has lived in Canada’s B3K, a kilometer south of the New Bridge. He finds the ambience of small urban Maritime Canada much like that of suburban Boston in the ’60s, and is most content with his housing situation.

McLeod’s recently completed career in Information Technology followed work as a laboratory technologist and some university mathematics education. He has done independent study in patterns and classification systems, achieving results in non-standard chemical nomenclature. Lately, these same researches have been extended into Jungian function theory and Myers-Briggs typology, so it would actually be fair to characterize him as an amateur psychologist.

For the last three years, McLeod has been trying to apply his talent for recognizing large patterns to the analysis of emerging macro-economic trends in US mortgage finance.