icture a six-lane highway with roughly 110 million cars. The posted speed limit is 55 miles per hour, but there is not a police officer in sight. Since there have not been any major accidents in years, it is common practice to travel at 90 miles to 100 miles per hour, and insurance companies are lowering their premiums – often regardless of the state of the cars.
That describes the U.S. mortgage market from 2003 to 2006. The story ended exactly as you would imagine: a massive accident with severe repercussions not just for housing, but also for the financial system and the global economy.
Today, the six-lane highway has been reduced to three lanes, as origination capacity has been halved. One is a fast-pass lane for customers who have been sitting in traffic the past couple of years and are now being rewarded for good behavior with access to historically low mortgage rates: the HARP lane. (The Home Affordable Refinance Program helps homeowners refinance who are underwater or near-underwater but current on their mortgages.) But for everyone else, the speed limit has been reregulated to 35 miles per hour. There are police officers at every mile marker, and the insurance companies are charging much higher premiums.
Where do we go from here?
Despite fewer lanes on the mortgage highway, we believe the U.S. housing market has bottomed and is showing clear signs of a gradual and broadening recovery. The upward trajectory of housing prices should continue at a moderate pace. Over the past 100 years, housing has appreciated at roughly the rate of inflation. It is only in the past 10 years that housing has traded with substantial volatility due to leverage and “affordability” products. We believe the tailwinds are in place for an 8%–12% appreciation in housing over the next two years. Over the longer term, we expect a return to historical normal performance for housing relative to the rate of inflation.
We consider several dynamics in developing our outlook on housing: household formation, inventories, affordability and access to credit and lending.
The foundation of housing demand is household formation, which can occur via population growth, immigration or a reduction in the size of current households (people moving out of their existing households to start their own). Household formation averaged 1.3 million per year from 1997 to 2007 but declined precipitously during the financial crisis, hitting a low of 357,000. At the end of December 2012, the rate − around 973,000 − was still running more than 300,000 below the pre-crisis 10-year average for several reasons: deleveraging from the great recession, overleveraged college graduates, overbuilding during the housing frenzy and a decline in the homeownership rate, to name a few. We expect household formation to revert to its historical average within the next year or two as the economy recovers.
The homeownership rate, however, has declined from a peak of 69.2% in 2005 to a 17-year low of 65.3% today, and we believe the descent will continue. We see this as one of the secular headwinds for housing.
Turning to supply, inventories of single-family non-distressed existing homes, shown in Figure 1, currently sit at 11-year lows, equating to less than five months of supply at the current pace of purchases − a very optimistic backdrop. As supply continues to get absorbed, builder confidence has increased, as evidenced by the National Association of Home Builders Index reaching its highest levels since the middle of 2006, and momentum in housing starts is positive, which will have an explicit impact on GDP via residential fixed investment in construction and remodeling, as well as implicit knock-on effects. Housing, which was a drag on economic growth throughout the financial crisis, is now positively contributing to GDP.
Looking at real estate nationally tells an incomplete story, however. State-by-state differences can be very dramatic, as Figures 2 and 3 show. The stock of housing is not fungible: You cannot take demand for a Chicago apartment and meet it with a farmhouse in New Canaan, Connecticut. (As an aside, if anyone would like a home in New Canaan, please give us a call.)
Geographical idiosyncrasies, such as population and housing stock, are also part of the story. Figure 3 illustrates changes in housing stock relative to population differences from state to state in 2010 and 2011.
Finally, the inventory picture would not be complete without looking at homes that may make their way to market via foreclosure, real-estate-owned transactions or short sales with delinquent loans – the so-called shadow inventory, shown in Figure 4. We believe shadow inventory currently stands at just over five million, but not all of these mortgages will be liquidated; we expect the number of homes that ultimately reach the market to fall between three million and four million. While that is high, north of one million liquidations were absorbed in 2012, and the momentum behind inventory drawdown has been extremely positive. One of the main drivers has been the participation of investors, who are attracted to the high rates of return in the low interest rate environment. As housing continues to appreciate and the capitalization rate on real estate declines, it remains to be seen if investor appetite will persist. We believe it will, albeit at a slower rate than in the past 12 months to 18 months.
Figure 5 shows “housing affordability”: An index value of 100 signifies that a family earning the median income has exactly enough income to qualify for a mortgage on a median-priced home with a 20% down payment. Aside from earlier this year, the affordability index has never been higher.
There are two important assumptions in the index. The first is that a prospective home buyer is able to put down 20% of the purchase price. The average home cost around $224,000 in 2012, requiring a down payment of around $45,000. Compared to times of robust economic growth and low unemployment, today the average household may not have the money or may be less willing to put down a large sum. The second assumption is that the prospective borrower can obtain a mortgage at the current market rate. Given the tightening of lending standards since the crisis, that leap is not a given.
We would also point out that the ratio of new home prices relative to median household income remains above the mean of the past 35 years, as seen in Figure 6, which speaks to the reliance of housing on low mortgage rates.
Access to credit and lending
The ability to get a mortgage at today’s historically low rates remains an issue. Figure 7 shows the average FICO scores and loan-to-value ratios of closed and rejected loans in December 2012, as processed by Ellie Mae’s database, which encompasses 20% of all U.S. mortgage originations.
We would argue that tight credit is a positive, as lending standards will only become easier. Credit expansion has materialized in both the auto and credit card sectors and should slowly spill over into mortgages. Capacity constraints, higher capital requirements and more regulation will cause this transition to take time but mortgage credit should begin to loosen.
One of the main reasons banks have been unwilling to expand the mortgage credit box is the potential liability of making loans that may eventually become delinquent and default. The Dodd-Frank Act, which has created a “safe harbor” for lenders of “qualified mortgages” − those that it can be reasonably assumed will pay off − will create a safer system but, in our view, will slow the process of easing lending standards.
We remain constructive on the state of the housing market but recognize the road is far from smooth.
On balance, we believe the positives outweigh the negatives and look for housing to appreciate 8%–12% over the next two years. Housing should have positive influences on consumer confidence and labor mobility.
In terms of investment implications, we believe both agency and non-agency markets offer opportunities to generate excess returns, while active management should be able to add value to structural allocations. Agency mortgage securities offer liquid investments that can be traded against each other as well as against other liquid interest rate markets, specified mortgage pools and, less frequently, structured mortgage products.
Non-agency mortgages continue to offer the best risk-adjusted returns in the sector, but specific security selection will matter much more given their recent high returns. Compared to investing directly in real estate, which requires time to close, lawyers, insurance and transaction costs, non-agency mortgages offer similar returns without the friction. Pairing non-agency mortgages with agency mortgage-backed securities potentially provides an attractive return profile across a wide range of economic outcomes.