The jumbo mortgage market shrank during the housing downturn. Jumbo mortgages, defined as loans with an original balance exceeding $417,000, contracted disproportionally after the onset of the housing crisis with house prices plunging and the private mortgage market shattered. As the housing market begins to recover and the private mortgage market heals, Goldman Sachs examine the future prospect of the jumbo mortgage market.
Growth opportunities in the jumbo mortgages market are significant
Voss Capital is betting on a housing market boom
The Voss Value Fund was up 4.09% net for the second quarter, while the Voss Value Offshore Fund was up 3.93%. The Russell 2000 returned 25.42%, the Russell 2000 Value returned 18.24%, and the S&P 500 gained 20.54%. In July, the funds did much better with a return of 15.25% for the Voss Value Fund Read More
There are three reasons why we are optimistic about the future of the jumbo mortgage market. First, rising house prices imply more expensive homes and a higher demand for jumbo mortgages. Second, today’s tight mortgage lending standards affect jumbo mortgage borrowers less; future easing of lending standards is likely to benefit jumbo mortgage borrowers more. Lastly, the private sector will pick up a larger share of the pie as the government tries to reduce its footprint in the mortgage market.
Three reasons why non-agency jumbo mortgages will grow
The US jumbo mortgage market took a big hit during the housing downturn. According to the Mortgage Bankers Association (MBA), total mortgage originations fell by a quarter from 2007 to 2012. In comparison, the origination volume of mortgages larger than $417,000 halved during this period. The decline is particularly dramatic in the non-agency sector (Exhibit 1). The disproportional decline in jumbo mortgage originations had much to do with the 30% fall in national house prices. In this Mortgage Analyst, we assess the prospect of the jumbo mortgage market and conclude that the future of this market is bright.
To be clear, in this report, we define “jumbo” mortgages as loans with an original balance above $417,000. A more typical definition of jumbo mortgages relies on the agency conforming loan limit which varies over time and across locations.1 We choose to use a simple cutoff of $417,000 to avoid changes to the conforming loan limit affecting our analysis. By this definition, jumbo mortgages constitute 5.4% of all outstanding mortgages by loan count and 19% by loan balance at present.
Our optimism for the jumbo market, and the non-agency jumbo market in particular, comes from three sources. First, we expect continued house price appreciation over the next few years, which will lead to an increasing number of expensive homes that may need jumbo mortgage financing. Second, we expect mortgage lending standards to ease among high-income borrowers with solid credit quality. Jumbo mortgage borrowers tend to fall into this category and may increase their borrowing on the back of improving credit availability. Third, currently, 20% of jumbo mortgages are guaranteed by GSEs and FHA/VA. As GSEs and FHA/VA contemplate reducing their dominant presence in the market, a growing share of the jumbo mortgage market will be picked up by the private sector.
The number of high-value homes is growing
The most important determinant of the size of the jumbo mortgages are the number of high-value residential properties. Assuming a standard 20% down payment, the $417,000 cutoff implies a property value threshold of $521,250. More properties valued above the $521,250 threshold suggest a higher demand for jumbo mortgages.
The geographical distribution of high-value properties is uneven. According to the 2011 American Community Survey (ACS), 9.9% of owner-occupied homes in the US were worth $500,000 or more. In Hawaii, Washington DC, and California, more than 30% of the homes were high-value properties. In West Virginia, Iowa, South Dakota, Mississippi, and Indiana, less than 2% of the homes were worth $500,000 (Exhibit 2).
Continued house price appreciation means that an increasing number of homes will become high-value properties. The Zillow Home Value Index shows that national house prices increased 5.5% in 2012. The Zillow Home Price Expectation Survey indicates that market participants expect house prices to increase 5.3% in 2013, 4.5% in 2014, 4.0% in 2015, and 3.5% in 2016. Applying these house price growth rates to the 2011 ACS propertylevel data, we find that the percent of homes worth $500,000 or more will increase from 10.4% in 2012 to 12.8% in 2016 (Exhibit 3). This means that there will be an additional 2 million homes worth $500,000 or more in 2016 compared with 2012.
Further Reading Student Loans Won’t Kill Housing Recovery
Lending standards are likely to ease for high-income borrowers
A rising number of high-value homes is one necessary condition for a growing jumbo mortgage market; another necessary condition is banks’ willingness to lend to the borrowers. Mortgage lending standards tightened significantly after the housing crisis. But high-income borrowers purchasing expensive homes are less affected than other borrowers for two reasons.
First, the tightening of mortgage lending standards was concentrated at the lower end of the credit quality spectrum. Exhibit 4 contrasts the distribution of borrower FICO scores among loans originated in 2002 and 2012. The difference in the patterns suggests that borrowers with impaired credit history face much more challenges in obtaining loans today than other borrowers: only 3% of the loans originated in 2012 had FICO scores lower than 640, compared with 14% in 2002.
Second, borrowers with high-value homes tend to have higher income and stronger household balance sheet than other borrowers. As a result, these borrowers are less likely to experience significant difficulties in obtaining loans even in the post-crisis environment of tight lending standards. The 2010 Survey of Consumer Finances (SCF) data suggest that borrowers with home value above $500,000 have more education (83% vs. 42% with a college degree), higher income ($178,905 vs. $63,023 median household income), more assets ($594,500 vs. $30,200 median financial assets), and lower debt burden (19% vs. 25% median back-end debt-to-income ratio) compared with other borrowers (Exhibit 5).
Going forward, lending standards are likely to loosen more for high-income borrowers with good credit quality. In fact, the April 2013 Senior Loan Officer Opinion Survey (SLOOS) conducted by the Federal Reserve reported that, relative to a year ago, a modest net fraction of banks were more likely to approve an application with a FICO score of 720 and a 20% down payment. However, about one-third of respondents indicated that they were less likely to approve FHA-insured home-purchase loans with FICO scores of 580-620 and a 3.5% down payment. The SLOOS data suggest that lending standards are indeed easing at the higher end of the credit quality spectrum, but continue to tighten at the lower end.
The share of GSE and FHA/VA loans is likely to shrink
Jumbo mortgage originations shifted dramatically from the private sector to the GSEs and FHA/VA after the onset of the housing crisis. In 2007, the vast majority of jumbo mortgages were originated through whole loans and private-label securities: the combined share of GSEs and FHA/VA was less than 5%. In 2008, the private-label securitization channel completely shut down, and whole loan originations fell precipitously. In 2009, two thirds of jumbo mortgage originations were backed by the government. Since then, non-agency jumbo originations have increased but remained low relative to pre-cycle levels. Exhibit 6 highlights the distinct patterns before and after the housing downturn: GSE and FHA/VA loans accounted for less than 2% of outstanding jumbo mortgages in 2007 and 20% in 2012.
The role of the GSEs and FHA/VA in the jumbo mortgage market will shrink in the future. In the Federal Housing Finance Agency’s (FHFA) Strategic Plan for fiscal years 2013-2017, the regulator of the GSEs stated that “FHFA will take steps to establish a path for shifting mortgage credit risk from the Enterprises to the private sector and gradually reduce the Enterprises’ proportion of the market.” For jumbo mortgages, continuing to raise guarantee fees and/or reducing the conforming loan limit will shift origination volumes from the GSEs to the private sector.
Similarly, in an effort to strengthen the Mutual Mortgage Insurance Fund, FHA has also been raising its mortgage insurance premium. Rising mortgage insurance premium affects jumbo mortgages disproportionally because larger loans are more sensitive to changes in interest rates than smaller loans. For example, larger loans tend to prepay faster than smaller loans when interest rates fall. Therefore, increases in the guarantee fees and mortgage insurance premium will induce substitutions from the government sector to the private sector. Put differently, even if the total size of the jumbo mortgage market remains the same, an increasing slice of the pie will go to the private sector as the GSEs and FHA/VA try to reduce their role in the future housing finance system.
In summary, we see growth opportunities in the non-agency jumbo mortgage market for three reasons. First, sustained increases in house prices will push more homes into the range requiring jumbo mortgages for financing. Second, potential borrowers of jumbo mortgages are less likely to be constrained by today’s tight mortgage lending standards.
Future easing of lending standards is also likely to benefit borrowers of high-quality jumbo mortgages more than other borrowers. Lastly, as the GSEs and FHA/VA contemplate reducing their role in the mortgage market, the private sector will pick up a larger proportion of the jumbo mortgage market in the future.
To be clear, the absolute origination volume of jumbo mortgages may not increase significantly going forward since we expect interest rates to rise and refinancing volumes to fall. The relative growth prospect of jumbo mortgages, however, is bright for the reasons discussed above. What are the downside risks that could prevent the non-agency jumbo mortgage market from growing considerably? We see three main risks.
First, the timing and scale of the return of private-label securitization remains uncertain. At the peak of the housing market in 2005 and 2006, annual originations of private-label securitized mortgages with original balance above $417,000 were around $450 billion. In 2012, less than $5 billion private-label securitized jumbo mortgages were originated. Regulations affecting private-label securitization, including risk retention and premium recapture, are still being debated, and these rules need to be clarified and other hurdles need to be removed before we see significant growth in this market.
Second, the appetite for holding 30-year fixed-rate whole loans among banks may be limited when interest rates begin to rise and the yield curve flattens. If whole loans on banks’ balance sheet cannot grow substantially and private-label securitization remains dormant in the coming years, the non-agency jumbo mortgage market will face challenges in meeting demand from potential borrowers.
Lastly, political pressure may induce the GSEs to raise the conforming loan limit as house prices increase across the country. Our baseline expectation is for the agencies to continue ramping up the guarantee fees and keeping the conforming loan limit unchanged or even reduced in order to decrease their market share. However, this is far from certain. If the guarantee fees stay low and the conforming loan limit keeps pace with house price appreciation, the private sector may not see significant growth despite increases in the overall size of the jumbo mortgage market.
Further Reading Student Loans Won’t Kill Housing Recovery