The GSEs: Good Intentions, Bad Results

“What we know about the global financial crisis is that we don’t know very much.”
Paul A. Samuelson, 1970 Nobel Economics Prize Winner

Nationalizing the Single Family Mortgage Market

Almost lost now in the shadow of more recent news, the US government effectively nationalized the Federal National Mortgage Association (Fannie Mae, FNM) and the Federal Home Loan Mortgage Corporation (Freddie Mac, FRE) on September 7, 2008 by placing both into conservatorship.1 Combined, they had either bought outright, securitized and sold to others, or guaranteed single and multi-family mortgages summing to over $5 trillion, representing over 40% of the $12 trillion of all such mortgages outstanding. This is immense: the entire United States public debt is “only” $9.5 trillion currently.

In nationalizing these government sponsored enterprises (GSE), Treasury Secretary Paulson essentially admitted that the combination of inadequate regulatory supervision, risk-taking encouraged by the stockholder ownership form, and innovative financing unaccompanied by effective risk management would lead to the shrinkage and possible eventual extinction of the GSEs.

The Treasury plan will limit the size of each GSE mortgage portfolio to $850 billion as of year-end 2009. After 2009, the mortgage portfolios for each GSE will shrink by 10% per year until the mortgage asset total falls to $250 billion. As a consequence, in the future, the GSEs are very unlikely to be able to supply substantial liquidity to the US mortgage markets or act as the marginal buyer.

How did we come to this unfortunate end result after such a long, mostly stable history for the GSE form of organization?

Sponsoring Home Ownership through Fannie and Freddie: Social Engineering Does Not Always Produce Intended Results

Fannie Mae had been originally chartered in 1938, but became a stockholder-owned GSE through Congressional action 30 years later. The intent was to provide funding for longer-maturity mortgages not requiring bullet payments and to allow for lower down payment percentages. In 1970, Congress created Freddie Mac to provide more competition to Fannie Mae in the secondary mortgage market, which involves the structuring and sale of securities collateralized by the mortgage loans or whole loan purchases. Both GSEs had been created by Congress initially to increase funds available to finance home purchases and later to make housing more affordable and accessible for low and moderate-income income cohorts. Later, each executive administration could through regulations and interpretations of Congressional laws prioritize its own set of housing objectives through the GSEs. As an example, in 1993 the Clinton administration issued new regulations under the Community Redevelopment Act that provided for more access to mortgage credit for inner city and rural communities. Many of these mortgages, which may not have been funded under previously established underwriting guidelines, involved borrowers with lower credit scores, smaller percentage down-payments, and lower debt service coverage ratios.

Although the Federal government did not legally have the obligation to guarantee the outstanding debt of the GSEs, investors had always assumed an implied federal guarantee to the GSE obligations. This allowed the GSEs to issue debt at lower interest rates than otherwise, to be exempt from state and local income taxes, and to use higher leverage ratios (i.e., invested assets to capital) than commercial banks . Over time, the GSEs were allowed to increase their leverage ratios to as much as 70:1, ($70 in mortgage assets could be supported with as little as $1 in capital). By the end of June of this year, Fannie Mae’s leverage ratio was 20:1, while Freddie Mac’s leverage ratio was 70:1. With such high leverage, asset impairment is magnified in a downtown, causing rapid capital depletion. It was this lack of adequate capital and the apparent difficulty of raising more capital that forced the Treasury’s hand.

Higher Home Ownership Rates, Creative Financing and High Leverage Lead to a Cycle Downturn

The home ownership rate2, which had moved in a rather narrow range from 64% to 66% from the late 1960’s to the early 1990’s, had increased to 69% by 2005. Average national home prices had increased by an annualized 11.3% from 2000 through 20063 while average income enjoyed only a modest increase. Normally, economic theory would predict that the higher the price of a good, the lower the demand. But real estate proved otherwise as more borrowers, through creative and expansive financing techniques, qualified for home purchases from the mid-1990’s through 2006.

The Federal Reserve Bank of Atlanta in Sept. 20074 estimated that more extensive financing alternatives, such as adjustable and hybrid rate mortgages, contributed from 55% to 70% of the increase in the home ownership rate. During the latter period of this mortgage expansion phase, from 2004 to 2007, Moody’s estimated that financial institutions and independent mortgage brokers made approximately 15 million mortgage loans that were poorly underwritten and made the dire prediction that up to 10 million of these loans will eventually default.

Meanwhile, the share of subprime mortgages to total originations rapidly increased from 5% in 1994 to 20% in 2006, representing approximately $600 billion, while the average rate difference between subprime and prime mortgages declined from an average of 2.8% in 2001 to only 1.3% in 2007. This rate differential occurred even while subprime borrower and mortgage loan risk profiles deteriorated.

Chasing Market Share and Accepting Higher Risk

During this phase of a substantial expansion in mortgage originations, the GSEs actually lost market share to private label originators. The percentage of originated conventional5 loans that were purchased by the GSEs declined from the 40% range in the years 2001 to 2003 to under 20% for 2006.

In seeking to reverse their falling market share of originations, the GSEs began to underwrite mortgages that have been characterized variously as “Alternative A”, “Expanded Approval” or “A Minus” mortgages. These mortgages were ultimately similar in credit quality to the older definitions of “sub-prime” mortgages. The GSEs also began to accept some “low-documentation” mortgages that had lower loan-to-value ratios than private label issuance and involved lower loan amounts.

In seeking to maintain mortgage securitization market share, the GSEs began to concentrate more on their relationships with larger mortgage originators, assisting in the development of the large “aggregator” model where underwriting partners, such as Countrywide Credit and Washington Mutual, provided ever larger percentages of the originated mortgages. As these large aggregators turned more to low documentation, alternative A, and interest-only adjustable rate mortgages (ARMs), the purchase wherewithal of the GSEs lent support to the growth of these alternative mortgage markets. The GSEs in turn appreciated the scale of mortgage operations and servicing capabilities of the large aggregators.

Starting in 2004, the GSE portfolios of subprime and Alternative-A loans and securities began to grow rapidly with the GSEs becoming the largest purchasers of such mortgages between 2004 and 2007. Eventually, the total GSE exposures to Alternative A and ARM mortgages from direct holdings and mortgage guarantees came to represent about 40% of the value of these mortgages. Ultimately, this concentration of loan origination risk and the loosening of underwriting standards caused sufficient loan loss write-offs and loss reserve accumulations to render the GSEs insufficiently capitalized.

Future Roles for the GSEs: Costs of Providing Mortgage Market Liquidity

“For years governments have been promising more than they can deliver, and delivering more than they can afford.”

Paul Martin, Canadian politician

The net costs to the Treasury from the credit crisis, with the mortgage market at the epicenter, may ultimately total in hundreds of billions of dollars. Through committing initially up to $100 billion in capital to each of the GSEs, the Treasury is on the hook for any ultimate mortgage losses in excess of shareholders’ equity, which could indeed prove to be negative on a mark-to-market basis for each GSE. The net cost of the recently enacted Emergency Economic Stabilization Act, if any, would be in addition to this.

With little capacity left to purchase mortgages due to the $850 billion mortgage exposure limitation, the GSEs are not well positioned to fulfill their traditional roles of supplying liquidity to the mortgage marketplace. As financial institutions have suffered approximately $600 billion in asset write-downs as of the date of this article, they are largely reluctant to commit capital to the mortgage markets and, therefore, do not provide a substitute for the GSEs in terms of funding mortgages.

It appears that the next Congress is likely to decide on the organizational form, operational scale and strategic mission for the GSEs or their successors. Enabling socially-oriented objectives, but with a lax regulatory structure and ineffective constraints on the business practices of the GSEs ultimately proved very expensive to US taxpayers.

Policy is always designed, of course, to improve overall societal welfare: implementation can occasionally cause a disconnect from intent. The next version of these structures will, we hope, have learned from these errors

Rich Palmer

Next Article