Never have the government-sponsored enterprises (GSEs) dominated the U.S. mortgage market as they do today. Through June 2012, the GSEs [Fannie Mae (FNMA.OB) and Freddie Mac (FMCC.OB) - and also including Ginnie Mae accounted for 100% of the mortgage market. (Note that this excludes mortgages originated by banks and retained on balance sheet.) It is remarkable to consider that the government's 95% share of the mortgage market in 2008 has turned out to be the low-water mark of the past five years. Rather than a temporary phenomenon in the wake of the worst housing crisis since the Great Depression, GSE dominance of mortgage finance continues to grow in strength.
The great irony is that this is occurring as officials in Congress and the Administration seem to agree that GSEs should be wound down. The problem is that winding them down requires a comprehensive plan for mortgage finance reform that allows for a seamless transition from the current regime so as to avoid any disruption to the nascent housing recovery. To date, the Administration has offered little more than brainstorming sessions, which makes it more likely that GSE dominance could continue well into the middle of this decade.
Perhaps the least appreciated aspect of the GSEs' mortgage market dominance is its likely impact on Fannie's and Freddie's cash flow relationship with the Federal Government. As explained by e21 in a 2010 editorial, "A Closer Look at GSE Credit Losses," Fannie's and Freddie's losses stemmed primarily from buying mortgages in overheated housing markets at the height of the bubble. Despite the popular focus on "subprime" loans and securities purchased since 2001, the bulk of losses came on interest-only and low-documentation loans made to higher income borrowers in Nevada, Arizona, Florida, and California, in 2006 and 2007. Moreover, the allegations that GSE losses came from banks "dumping" bad loans on the GSEs after conservatorship never passed serious scrutiny. The loans acquired since the government's takeover were of significantly higher quality than those purchased in the years prior to the bailout.
According to the most recent FHFA data, Alt-A and interest only mortgages have combined for nearly 70% of all credit losses since 2008. Likewise, loans in the four bubble states (Arizona, California, Nevada, and Florida) have contributed 57% of losses, despite accounting for just 25% of all loans. Finally, loans originated in 2006 and 2007 account for 64% of all losses despite comprising just one-third of the portfolio. These three categories all represent basically the same loans - think of an interest-only loan made to a borrower in Nevada in 2006, for example. These 2006 and 2007 vintage mortgages are the bulge in the belly of the snake. Once they've been digested, the outlook for GSEs' finances is extremely favorable.
Between 2008 and 2011, Fannie and Freddie recorded $91 billion in credit losses (charge-offs) on mortgages. In 2012, this figure has fallen to just $14 billion. More significantly, the amount of expected future losses has totally collapsed. Between 2008 and 2011, the GSEs had to generate a combined $203 billion in reserves against likely future losses. These losses (writedowns on the value of loans) were the primary driver of the GSE bailout.
In 2012, the combined provision against future credit losses was just $2 billion and Fannie Mae actually withdrew $1 billion from reserves because expected future losses have declined even further. Loans made since 2008 have higher credit scores, larger down payments, and substantially lower default rates. Most significantly, loans made since house prices have bottomed are unlikely to perform poorly since being underwater (a home worth less than the unpaid mortgage principal balance) is generally consider a necessary condition of default.
In addition to better credit performance, there are several other factors at work that will likely boost revenue and profits on new loans for the GSEs. In fact, Fannie and Freddie are likely to be more profitable than even the 2000-2005 period because they: 1) face less competition today; 2) are moving g-fees higher (at the bequest of FHFA and Congress); 3) are allocating less for compensation-related expenses (e.g. some operating earnings in the past went to extravagant bonuses); and 4) overall margins are up because of more conservative underwriting (e.g. higher FICO scores and lower loan-to-value ratios).
There is one more profit driver that's potentially even more important - and it's almost getting no public attention. In the second quarter of 2012, Fannie Mae's net income was $5.1 billion. The same quarter, Freddie Mac earned $3 billion. While the GSEs are currently earning profits at a $32 billion annual rate, the accounting profit could actually increase substantially as previous losses from 2008-2011 are "carried forward" to offset future tax liability. The value of these "deferred tax assets" are counted as assets on the balance sheet and accounted for as net income in the quarter in which they're recognized. In 2008, Fannie and Freddie established a "valuation allowance" of $35 billion against previously accumulated deferred tax assets. Were these gradually added back to the balance sheet due to the current outlook, combined accounting profits could exceed $60 billion in 2013.
In August, the Obama Administration responded to the positive outlook by amending the terms of the bailout to require the GSEs to distribute all profits to the Treasury as dividends each quarter. This decision was partially rationalized by the desire to end the circularity of having the GSEs borrow money from Treasury only to then turn around and pay the 10% dividend rate on past bailout funds. However, this rationalization for the change makes little sense. When the GSEs are profitable, as they are today, they can fund the dividend payments out of current income. There is no need to borrow from Treasury. If there was concern about the circularity, it should have been addressed in 2009-2011 when the GSEs borrowed $35 billion from Treasury to pay the Treasury $35 billion in dividends. Today, the cash flows are one-way (from the GSEs to Treasury). If there is concern that profits won't be sufficient to meet the $18 billion in annual dividends owed Treasury, a better amendment might have been to allow the GSEs to repay taxpayers by repurchasing senior preferred shares with any retained earnings.
Instead, the Obama Administration's amendment is likely motivated by a desire to translate GSE dominance of the mortgage market into increased federal revenue. For example, in the second quarter of 2012, the GSEs paid Treasury $5 billion in dividends and retained the other $3 billion in profits. This was the first time the GSEs were profitable enough to generate retained earnings in excess of the dividend payment to Treasury. In the future, the Treasury will keep the entire $8 billion.
This creates an interesting dilemma: if the Treasury has a direct financial interest in the GSEs' profitability, what incentive is there to reform the mortgage market in a manner that would reduce federal revenues? Rather than "buying time," the Obama Administration's move may actually be cementing the current arrangement and increasing the odds that the 95% market share from 2008 remains the GSEs' low water mark for the next 10 years as well.