Last weekend the Federal Housing Finance Agency announced the first in a likely series of changes at government-sponsored entities Fannie Mae and Freddie Mac.
This first one is sensible, overdue, and perhaps helpful to mortgage rates: the rollout of a uniform mortgage-backed security, as opposed to ones specific to Fannie or Freddie. Uniformity of any security makes it more liquid and cheaper to trade.
The prospect of larger and harmful changes appeared with the April confirmation of Mark Calabria as director of the FHFA by a 52-44 party-line vote in the Senate. Calabria has been a critic of the mortgage agencies and strong advocate of privatization. Together with the Treasury, FHFA will announce a privatization plan this month. Brace yourselves.
Before discussing the risks of privatizing, some background on the GSEs and mortgage market as a whole will help.
Today there are about $10.3 trillion in U.S. first mortgages outstanding, only a little larger than in 2008 after the period of bubble growth (Fed Z.1). For scale, the entire sum of U.S. bank credit (Fed H.8) is only $13.4 trillion, which includes about $2.6 trillion in the overall total of mortgages.
The rest of the $10.3 trillion after the bank portion is roughly $1.2 trillion insured by the Federal Housing Administration, and another $0.6 trillion guaranteed by the VA. Another $1 trillion sits in credit unions, private MBS (the worst of the worst in the bubble, down from $2.2 trillion foreclosed to $0.4 trillion), REITs and miscellaneous lenders.
Which leaves roughly $5 trillion owned or guaranteed by Fannie and Freddie. Right there is the deep story. The GSEs got in to trouble way back in the late 1970s by jumping charter as facilitators and began to buy, hold, and finance huge quantities of mortgages with very little capital for the benefit of their stockholders. By 2008 the owned and financed $1.7 trillion (the “retained portfolio”). When panic struck the world in 2008 the GSEs were unable to roll over the debt financing the portfolio — that’s why and when they were placed in conservatorship by the FHFA, where they remain.
Since 2008 the GSEs have shrunk their holdings by 75 percent and paid down debt correspondingly. They operate today really as insurance companies, taking fees to guarantee loans. They “buy” loans only as sidings on the way to markets and to replenish small continuing holdings as they pay off.
Those who would privatize the GSEs have a variety of motivations. Some the well-intended faith in markets and opposition to government meddling. Others want lending to be bank-only. However, the primary motivation is to reduce taxpayer risk, which is badly misunderstood.
What is the real risk? And considering that GSE risk is a powerful benefit to taxpayers — $5 trillion in cheap and reliable mortgages ain’t hay.
GSE risk, from mortgages in general, is routinely propagandized. The VA mortgage guarantee dates to 1944, and these loans have never required a down payment — and the VA has never required a subsidy or bailout. The FHA mortgage insurance program dates to 1938 and likewise has never required a subsidy. It has gotten close and may need help someday, which is reasonable considering its role of assisting credit to the weakest of borrowers. These loans are commonly securitized by the Government National Mortgage Association (Ginnie Mae, 1968), its MBS guaranteed by the Treasury and never losing a dime.
In the great 2008 meltdown, the largest bank run of all time, investors ran on Fannie and Freddie. The GSEs survived by drawing $187 billion from the Treasury, then paid it all back in five years, and have surrendered another $110 billion in earnings — essentially all — since. That public hears “bailout,” not temporary assistance quickly repaid and then some.
As risk goes, the GSEs do not keep me up at night. Fiddling with them for the sake of privatizing philosophy scares me.
Private MBS did the big damage in the bubble. Yes, better not to have big government agencies, but it’s a lot easier to keep track of the fox if penned in your yard than loose.
The risk of the GSEs is more than offset by other benefits. Uniformity has more benefit than just MBS: everyone understands GSE underwriting requirements.
And those constantly refined requirements work: They are automated into excellent software, and post-bubble mortgage delinquency is the lowest measured.
Nearly every mortgage, GSE or not, is closed on GSE promissory notes and deeds of trust. Since the product is uniform, competition among lenders is fierce and the product transparent to consumers.
Consider the one mortgage market privatized since 2008: Jumbos. There are at least 15 unique wholesale providers in volume, just the fixed-rate portion (there are too many adjustable-rate providers to count, several thousand).
Each has a different pricing model and different underwriting guidelines, each set is several dozen pages, and each is plagued by capricious box-checking unhinged from actual credit risk. The jumbo market is totally opaque to consumers, many of whom (and exasperated loan officers) pursue a cheap rate only to find that some quirk in financial profile excludes them.
Keep a sharp eye on Calabria. He is a true believer in privatization. The odds are against the June plan, no matter what it is, it will be likely to raise the cost of borrowing or exclude segments from credit. Calabria has already pointed out that he can end the GSE conservatorship by FHA power “without Congress.”
