The 45% drop in earnings at Freddie Mac, for a quarter that ended in June, predating August’s intensification of the mortgage market crisis, indicates an uncomfortable reality. President George W. Bush may attempt to hold back the tide of foreclosures by handing out yet more federal guarantees to subprime borrowers, but that will only delay the inevitable. Far from being able to assist the mortgage market by purchasing yet more mortgage backed securities, Fannie Mae and Freddie Mac are about to fight for their lives. The more interesting question is what should be done about it if they perish.
Fannie Mae and Freddie Mac were set up in 1968 and 1970 to extend the bizarre US policy of having third party government-related entities guarantee home loans. Other countries don’t do this; Britain had for many years a perfectly satisfactory system of building societies, mostly mutually owned, that made floating rate home loans to approximately the same percentage of the population as were homeowners in the United States. While the British system has gone astray in recent years, it remains blessedly free of government bailouts though not of overpriced housing – in London it is prices that have reached truly gargantuan proportions, rather than the houses themselves as in the US.
Fannie Mae and Freddie Mac can get in financial trouble not only from accounting incompetence, though their inability to determine within a billion or two what their true profits are is fairly startling, but through market movements. In 1981-82 Fannie Mae made losses — $190 million in 1981 and $105 million in 1982. Those losses were made on a base of total outstanding guaranteed mortgages of $62.1 billion at December 1981, so represented 0.31% and 0.17% of outstanding assets respectively. Based on today’s asset book (at 31st December 2006) of $2,526 billion, Fannie would in a similarly problematic environment make losses totaling $12 billion, compared with its capital of $41.5 billion.
Fannie’s defenders would claim that its derivatives portfolio would prevent such losses in today’s market, but relying on Fannie’s derivatives trading skills seems optimistic. The interest rate risk on a portfolio of home mortgages is almost impossible to hedge, because prepayment rates vary wildly depending on interest rate movements and housing market conditions. All you can do is observe trends that have held good in the past and hedge against a repeat of them. However, because of the chaotic nature of any large economy, there is no guarantee that such trends will manifest themselves in the future. It is thus more than possible that Fannie Mae’s derivatives portfolio, in an extreme situation, would magnify the institution’s losses rather than reducing them.
There is a further problem. The 1981 and 1982 losses occurred in a period of intense interest rate volatility; they did not however coincide with a significant house price decline. Because inflation was so rapid during those years—12.5% in 1980 and 8.9% in 1981, though only 3.8% in 1982 – nominal house prices dropped very little, except in pockets of industrial devastation such as Youngstown, Ohio. Fannie Mae’s credit losses were thus modest. This time around, it is already clear that house prices are dropping – the Case-Shiller index was down 3.4% in the second quarter of 2007 compared to 2006 – and with mortgage markets tightening, they are very likely to drop further, perhaps to a final low 15% below the peak nationwide and 30% below it on the coasts, as this column forecast a year ago.
Needless to say, a 15% drop in house prices would spread mortgage losses far beyond the subprime sector, to which Fannie Mae and Freddie Mac have less than average exposure. With a total mortgage book of $2,526 billion (on and off its balance sheet) Fannie Mae would only have to incur loan losses averaging 1.64% of principal to reach insolvency. While many of its loans are sufficiently “seasoned” to have benefited from several years of home price appreciation, the epidemic of mortgage refinancing in 2002-05 will have made that cushion much smaller than it might be. You thus don’t need to imagine the Great Depression, merely an extended period of poor housing markets, to postulate a scenario in which Fannie and Freddie expire (or more likely are bailed out by the U.S. taxpayer.)
If such a bailout becomes necessary, it is to be hoped that Congress will look once again at the fundamental question: what should government’s role be in the home mortgage market, and what sort of mortgage market do we really want?
The traditional U.S. home mortgage market at its best is epitomized by the Jimmy Stewart 1946 classic movie “It’s a Wonderful Life”. Home mortgages are made by savings and loans, which are purely local organizations. In order to get a mortgage, a borrower needs to have an account with the S&L, with a substantial amount on deposit. Jimmy Stewart, not a wealthy man, owns the local S&L, knows most of the borrowers personally, and has played a huge role in the development of his local town of Bedford Falls.
All very sweet and sentimental, you may respond, but today’s mortgage market has greatly increased efficiency, removing the barriers between borrowers and lenders, creating a liquid worldwide market in which German state banks and the People’s Bank of China can alike invest, and thereby hugely reduced the borrowing costs of the American homeowner. Isn’t that right?
No, actually it isn’t. For the earliest month for which the Federal Reserve has mortgage yield data, April 1971, the 20-year Treasury bond (there being no 30 year bonds then) yielded 6.00%. 30 year fixed rate conventional mortgages yielded 7.31%, a differential of 131 basis points (1.31%). In the first month of data for the 30-year bond, April 1977, it yielded 7.75% while 30-year conventional fixed rate home mortgages yielded 8.67%, a differential of 92 basis points. Fannie and Freddie had only just got going in 1971 and were young in 1977; it’s a pity pre-Fannie 1950s data is not readily available, but I’m sure it would show the same general level of differentials.
And today? In July 2007, the latest month for which data is available, before the latest unpleasantness really got going, 30 year Treasuries yielded 5.11% and 30-year conventional mortgages 6.70%, a differential of 159 basis points. For February 2006, at the height of the housing boom, 30 year Treasuries yielded 4.54% and 30-year conventional fixed rate mortgages 6.25%, a differential of 169 basis points.
Averaging each pair of observations, a typical yield premium over long dated Treasuries for 30 year home mortgages was 112 basis points in the 1970s and 164 basis points today. Effectively, government guarantees, securitization, fantastic modern analytic technology and the combined genius of Wall Street have made home mortgages 52 basis points, just over ½% per annum, more expensive.
Anyone who has read the newspapers over the last few years can begin to guess what the problem is. Wall Street is full of centi-millionaires and even billionaires, while the wealthier mortgage brokers also drive Lexuses. House prices in real terms are far higher than 30 years ago, and mortgages are made on an entirely impersonal basis, using computerized credit scoring, with everybody skimming off commissions and nobody responsible for the credit risk until the mortgage becomes a tiny part of the assets of an offshore securitization conduit of a dozy German bank. The above analysis shows that these people are not economically adding value, but are mere successful rent-seekers.
Alert readers will at this point ask: if the new system is so much less efficient in delivering value to the ultimate customer, how did it drive out the old?
The answer is one word: salesmanship. Anybody who has lived in a US suburb with an economically attractive zipcode and no butler will recognize that excessive salesmanship is the bane of American life. This is even more true in the mortgage business. Homeowners today don’t go into their local S&L, save for half a decade and request a mortgage from Jimmy Stewart. Instead they are sold a mortgage product, either directly or over the Internet, by an aggressive salesman who is paid a multiple of what Jimmy Stewart earned, or at least aspires to be. That product is then securitized by an investment bank trader who in good years is paid a LARGE multiple of what Jimmy Stewart earned. With others it is sold to a securitization vehicle of immense complexity that has been set up by Wall Street lawyers paid a HUGE multiple of what Jimmy Stewart earned. Costs have been increased at every point in the process, but aggressive salesmanship has driven Jimmy Stewart out of business.
The government has played a role too, primarily in the early stages of securitization development. Government guaranteed mortgages, originally for low income housing under the aegis of Ginnie Mae, were the first to be securitized, in 1968. It is likely that the mortgage bond market would never have achieved sufficient investor acceptance without its government guarantees – investors would have been rightly suspicious of a package of mortgages to unknown homeowners scattered around the country, and would have demanded a yield high enough to make the transaction impossible. However, it is likely that government guarantees (or those of the “government sponsored entities” Fannie Mae and Freddie Mac) are no longer necessary except perhaps in low-income housing where an element of subsidy is involved.
The Federal Trade Commission should do something about Fannie Mae and Freddie Mac’s advertising, and that of the mortgage banking industry, requiring them to run corrective campaigns with the slogan: “Interfering unnecessarily in your home mortgage to make it ½% more expensive.”
Skeptics may object that two technological advances at least may have made obsolete the Jimmy Stewart model of individual credit assessment and funding by personal savings. Those are credit scoring, by which computerized models are used to assess borrower credit quality, and money market funds, by which savings are bundled and placed through the wholesale money market. However mortgages are not credit cards; the amounts are much larger and the terms longer, so they would certainly repay an individual credit assessment, particularly if savings for down-payments were also required. Moreover, with computerized bank teller procedures and universal deposit insurance, S&Ls could today compete with money market funds at only a modest additional funding cost, and would use wholesale markets to smooth local gluts and scarcities of savings. Another new development, the interest rate swap market, can now minimize the mismatch between 30 year mortgage rates and 3 month deposit rates, which got the S&L industry in trouble in 1980-82.
If Fannie Mae and Freddie Mac go bust they should not be bailed out; they are unnecessary, as is the whole system of government guarantees. Instead legislation should establish a new system of home mortgage institutions, requiring a minimum 10% down-payment for mortgage loans and enjoying federally guaranteed deposit insurance. At the same time, a stamp duty of say ¼% should be imposed on securitization transfers, making that market illiquid and unattractive to Wall Street and investors. The home mortgage interest deduction could also usefully be repealed, reducing the attraction of McMansion ownership to the wealthy (and making little difference to those of modest means, whose mortgage interest payments generally fall within or close to the “standard” income tax deduction.)
No giant fortunes, but only moderate ones would be made from the new institutions. However they would reduce the cost of home mortgages and remove an entirely parasitic sector from the US economy. Because of the down-payment requirement, they would also perform the economically essential function of increasing the US savings rate.
Most important, such a system, by making “flipping” impossible and McMansion ownership more expensive would decrease house prices, bringing homeownership much more securely into the financial reach of those with moderate incomes. Homeownership would thereby increase rather than diminish overall, with only cheats, speculators and vulgar egomaniacs worse off.
It’s time to bring back Jimmy Stewart!
-0-
(The Bear’s Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that the proportion of “sell” recommendations put out by Wall Street houses remains far below that of “buy” recommendations. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)