In making my case for continued housing weakness, I’ve emphasized the negative effect of excess inventories on house sales, prices, new construction and just about every other aspect of residential real estate.
In housing, as in every goods-producing sector, excess inventories are the mortal enemy of prices. Lower prices are needed to unload surplus inventory, yet they also lead to the creation of more inventory by anxious sellers. The plight of house sellers and the reluctance of buyers are made worse by the realization that house prices can fall, and are falling for the first time in 70 years.
There are about 2 million excess housing units in the U.S., over and above normal inventory working levels. Before the housing collapse began in 2006, housing starts and completions were volatile but averaged about 1.5 million per year. So a 2 million excess is much more than the previous annual average build.
Furthermore, that excess is rising as homeownership declines as a result of foreclosures, unemployment, inability to meet mortgage standards or reluctance to own a depreciating asset.
Many people think that house inventories are coming under control. They point to the declines in inventories in relation to sales for new and existing homes, yet that calculus doesn’t include the 5 million or so housing units with delinquent mortgages or those in foreclosure, much less the additional troubled loans that are probable in years ahead.
They also don’t include foreclosed vacant houses that haven’t been listed for sale and vacant units that owners pulled off the market. These vacancies are included in the Census Bureau category called “Held off the market for other reasons,” and they now number 3.6 million, up 1 million from 2006. Falling house prices are associated with declining residential listings as disappointed sellers retreat in hopes of higher prices later.
-- Foreclosures Down: New foreclosures have dropped considerably in the last two years, but for temporary reasons. RealtyTrac Inc. estimates that there were 804,000 bank repossessions in 2011, down from 1.05 million in 2010. Nevertheless, the Federal Housing Administration’s seriously delinquent mortgages, often foreclosures in waiting, jumped from 8.2 percent of the loans it guaranteed in June 2011 to 9.6 percent in December.
The federal government encouraged lenders and mortgage servicers to delay foreclosures as modifications were attempted. There was also the voluntary moratorium on foreclosures during the robo-signing flap. This pause continued while the five largest mortgage servicers -- Ally Financial Inc., Bank of America Corp., Citigroup Inc., JPMorgan Chase & Co. and Wells Fargo & Co. -- worked on the recent settlement with the federal government and state attorneys general that called for $25 billion in mortgage writedowns and other aid to homeowners.
-- The Logjam Breaks: With that settlement completed, mortgage servicers and lenders will probably step up foreclosures. When they do, the so-called real estate owned -- the properties owned by lenders -- will be dumped on the market with all deliberate speed.
The effect on prices will be dramatic. The National Association of Realtors’ survey for December 2011 found that foreclosure sales were at an average price discount of 22 percent, compared with 20 percent in December 2010. Short sales, in which the lender forgives the difference between the sale price and the mortgage principal, closed 13 percent below market value. As of the second quarter of 2011, RealtyTrac found that real-estate-owned sales were at a huge 40 percent discount while short sale discounts averaged 12 percent.
These discounts tend to drag down the prices of other existing houses and force homebuilders to sell properties below cost in order to compete.
The trigger of renewed foreclosures will probably initiate another big drop in house prices, returning them to the long-term trend identified by Robert Shiller of Yale University. This measure of median single-family-house prices is adjusted for general inflation and for the tendency of houses to get bigger over time and therefore more expensive.
With these two corrections, prices in 1990 were about the same as they were a century earlier. Then came the bubble, followed by collapse, but it still will take a 22 percent decline to return prices to the flat long-term trend that prevailed between 1890 and 2000. Because corrections often overshoot on the downside, our forecast of a further 20 percent decline may be conservative. That would bring the total peak-to-trough decline to 46 percent.
-- Spreading Effects: That further drop would have devastating effects. The equity of the average homeowner with a mortgage has already dropped to 17 percent, from almost 50 percent in the early 1980s, due to home-equity withdrawal and falling prices. An additional 20 percent price decline would push homeowner equity into single digits with few borrowers having any appreciable equity left. It would also boost the percentage of mortgages that are underwater to 40 percent, from the current 22 percent, according to my calculations. The existing underwater loans have already created a $750 billion gap between mortgage principals and house values, according to CoreLogic Inc. The negative effects on consumer spending as well as mortgages and mortgage-backed derivatives would be substantial.
-- How Long?: A principal reason that median single-family-home prices are likely to fall an additional 20 percent is that it will take years to work through the excess house inventory, giving plenty of time for surplus units to depress values. I expect housing starts and completions, now about 650,000 at annual rates, to average 700,000 annually in future years. About 300,000 of those will replace housing units that are torn down or converted to other uses. So the net supply is about 400,000.
The demand side is determined by net household formation. Contrary to popular belief, household formation isn’t closely correlated with population, at least not in a cyclical time frame. By definition, a household is one or more people occupying a separate dwelling unit. So all the forces that make people want to rent or buy -- house prices, unemployment and mortgage standards -- play a role.
Household formation is about as volatile as housing starts and homeownership rates. It surged a decade ago when owning houses was the route to quick wealth; it dropped as prices collapsed. In the boom days when house prices increased 10 percent a year, a homeowner with a 5 percent down payment made a cool 200 percent on his investment each year, neglecting mortgage interest, maintenance and taxes. And, as a bonus, that person had a place to live rent-free.
-- Annual Absorption: Household formation in the fourth quarter of 2011 was 659,000 at annual rates. Over the last decade, it has averaged about 900,000, a number that seems reasonable in years ahead. Note, however, that this number may be on the high side if significant doubling up reduces household formation. Demand of 900,000 and net supply of 400,000 per year, as discussed earlier, will absorb 500,000 of the excess inventory annually. So the 2 million surplus of housing units I’ve identified will take four years to work off.
That would extend the bear market in housing to 2015, a full 10 years after it started.
One of the biggest contributors to this lengthy resolution is that Americans, armed with first-hand experience of falling prices, are beginning to separate their abodes and their investments. In the days when owners thought house values never fell, they bought the biggest homes they could finance. They now know otherwise. Further weakness in the prices of single-family houses and condos due to the depressing effects of excess inventories will add fuel to the fire.
Contrary to general belief, a single-family house, excluding the effects of increasing size and general inflation, has been a flat investment for more than a century. Such properties provide a place to live, but that value is offset, at least in part, by maintenance, taxes, utilities, real estate commissions and other costs. Furthermore, even with the tax deductibility of mortgage interest, renting a single-family house or apartment is cheaper than owning, absent price appreciation. This trend will accelerate in the years of deflation I foresee, when nominal house prices will probably fall on average.
The separation of abodes from investments should work to the advantage of rentals in future years, as it has since the housing bubble burst in 2006.
I’m not suggesting that Americans will give up on single-family owner-occupied housing. That ambition is too deeply embedded in our culture. But many will be more inclined to rent, including empty-nesters who decide to unload their suburban money pits, especially because their homes are falling in price.
Young couples may decide that because houses are no longer a great investment, there’s no reason to strain their financial, physical and emotional resources to buy big, expensive ones as soon as possible. They’ll stay in rental apartments a bit longer and wait until their children are of the age that a single-family house makes sense.
Retiring postwar baby boomers -- those who aren’t locked into underwater mortgages -- are also likely to rent as they separate their investments from their abodes.
-- Single or Multifamily?: I’ve made the case for about 4 million new renters in the next five years or so. But will they rent apartments or single-family houses?
Investors are buying foreclosed and other housing units, most of them single-family. Some did so in 2010, when the new homeowner tax credit briefly raised prices. They expected to flip them promptly, but instead became landlords as prices resumed their decline. Nevertheless, the interest of investors, who are often all-cash buyers, persists. The Realtors’ association reported that in December 2011, 31 percent of all existing house sales were for cash, 21 percent went to investors and 31 percent to first-time homebuyers.
For investors, however, managing single-family houses is challenging. An apartment usually has one or two walls exposed to the weather, but a single-family house has four plus a yard to maintain and a roof that can leak.
-- Apartment Building: Even as investors are buying single-family foreclosed houses, rising apartment rents and declining vacancies have spawned a miniboom in multifamily housing starts, albeit from close to a zero base. Furthermore, some of the growth may be in anticipation of demand from new retirees. Multifamily completions have yet to reflect this trend because it takes about 12 to 18 months to finish an apartment building.
Supply will probably continue to be augmented by conversions of unsold condos to rental apartments. Will multifamily housing soon become more overbuilt and end the increase in rentals and decrease in vacancy rates?
Those in the industry say that until recently, multifamily developers have been cautious. Tight financing has been one reason, with lenders providing 50 percent to 60 percent of the financing, compared with 80 percent in the salad days of 2006. Also, developers, accustomed to 8 percent to 10 percent capitalization rates, are reluctant to accept today’s returns of 4 percent to 5 percent. And banks are hesitant to lend to developers with bad records and favor borrowers with fortresslike balance sheets.
In addition. Fannie Mae and Freddie Mac have shied away from multifamily housing after getting stuck with bad apartment loans they bought in 2007 and 2008 as private lenders withdrew. The agencies’ share of multifamily loan purchases leaped to 85 percent in 2009, from 29 percent two years earlier. By 2009, they owned or guaranteed 40 percent of the $325 billion multifamily mortgage market.
-- Slow Start: According to Reis Inc., a New York-based commercial real estate research company, less than 40,000 new multifamily units were finished in 2011, the lowest number in 30 years. In 2012, Reis expects 72,000 to 85,000 new units. Even with robust apartment demand, net absorption -- the surplus of demand over new supply -- fell to 153,000 units in 2011 from 225,000 in 2010. In October of last year, the National Multi-Family Housing Council, a trade group, reported that its National Tightness Index was 56, down from 82 in July and a peak of 90 in April. A reading above 50 indicates that markets are tightening.
My industry contact indicates that conditions justify apartment-building in some markets, such as Los Angeles, San Francisco, New York, Boston, Chicago and Washington, where capitalization rates of about 5 percent for Class A buildings prevail. In many other areas -- such as Detroit and Cleveland -- rents don’t justify new construction and capitalization rates of 6 percent to 7 percent for existing apartment buildings are the rule.
On balance, lender caution will probably curtail any developer zeal to overbuild rental-apartment buildings for a number of years, at least in most cities.
-- Single-Family/Apartment Split: It’s difficult to estimate exactly how the 3.9 million net new renters I forecast through 2016 will be split between rental apartment dwellers and renters of surplus single-family homes, but I will venture some projections. The return of the rental vacancy rate to its earlier norm of 7.6 percent would provide about 750,000 units. An additional 1.5 million would result from an increase of multifamily starts and completions to the earlier norm of 300,000 per year, from the recent annual rates of 200,000. That would leave 1.7 million to be supplied from single-family rentals.
These projections are in line with my estimate that the 2 million excess inventories would be eliminated over the next four years. This consistency suggests that in four or five years, the housing market will return to normal, with the ownership rate back to its 64 percent norm and 3.9 million new rentals supplied by the elimination of excess inventories as well as the return of multifamily starts and completions to the earlier 300,000 annual rate.
Single-family starts and completions in this scenario would continue at current depressed levels of about 400,000 per year, but the elimination of 1.7 million single-family units in excess inventory by converting them to rentals would relieve the downward pressure on prices.
-- Only Projections: These numbers, however, reflect plausible but uncertain assumptions. A lower homeownership rate than the 64 percent average is possible now that Americans know house prices can and do fall. If so, more single-family houses would probably be converted to rentals and apartment construction could be stronger. If more people double up, household formation will be weaker and it will take longer to work off excess inventories, unless weaker new residential construction provides an offset.
In any event, excess house inventory, over and above normal working levels, will be gradually worked off over the next four or five years by new household formation. But about 4 million of the 4.5 million increase in households will be renters of apartments or rental single-family houses.
(A. Gary Shilling is president of A. Gary Shilling & Co. and author of “The Age of Deleveraging: Investment Strategies for a Decade of Slow Growth and Deflation.” The opinions expressed are his own. Read Part 1 of the series.)
This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.
To contact the author of this story:
A Gary Shilling at firstname.lastname@example.org
To contact the editor responsible for this story:
Max Berley at email@example.com