To the saying, "there are lies, damn lies and statistics," MIT economist William C. Wheaton might add another line: "and housing model forecasts."
In a packed IAP seminar Jan. 30 on "How Far Will the Housing Bubble Burst?", Wheaton, who holds a joint appointment in the Departments of Economics and Urban Studies and Planning, showed how forecasting models based on historic data may not match the complicated reality of today's housing market.
Wheaton's own econometric models of home pricing, based on data through 1998, were used to "forecast" housing prices from 1998 to 2005 in major American cities. The models "predicted" that housing prices would undergo a "correction" and fall as job growth and income stalled, particularly after the dot-com bust. Yet as Wheaton demonstrated through a series of graphs and charts that compared predictions with reality, in cities like Phoenix, Denver and Las Vegas, housing prices rose, sometimes spectacularly, in those years.
"The model says we should have seen (lower) housing prices. But in Phoenix, they went up 80 percent. Eighty percent!" Wheaton said.
Similar increases in Southern California, Florida and Washington seemed "totally unexplainable."
Wheaton, who helped organize the MIT Center for Real Estate, outlined two hypotheses that he says explain 65 percent of the model error.
One factor is an increase in the purchase of second homes. In the last two years, 2 million housing units were built, with only 1.25 million new household units, meaning a huge number of people are buying second homes, either as investments or residences. "People are buying housing that the economy, in some sense, doesn't need," he said.
The other factor is the emergence of the sub-prime lending market in the 1990s. Sub-prime loans, which charge high interest rates and are often held in a high-risk pool, are usually made to those with bad credit or who do not want to make a down payment. Such loans "push the envelope" of home ownership, Wheaton said.
From 1994 to 2005, the stock of sub-prime loans rose from 1 to 8 percent of the nation's total mortgage debt. Home ownership is rising, even though the national poverty rate is about 22 percent. "People technically in poverty are owning homes,'' Wheaton said. Since 1995, 5 million renters have become homeowners.
Now, if second-home buyers decide they prefer the stock market to real-estate investment and seek to unload their extra properties, that could be part of a "perfect storm" of factors that send housing prices tumbling, Wheaton said. But, second-home sellers are generally more willing to hold out for higher prices than those who need to sell a home because they are buying a new one, preventing prices from softening. However, "this is new territory for us real estate economists," Wheaton said.
More worrisome in the "perfect storm" scenario is what will happen to those first-time homeowners hit by variable rates on their sub-prime loans. What if the 5 million recent buyers were foreclosed upon? They would be forced back into the rental market. (Indeed, Wheaton speculates that some apartment owners are anticipating this.) However, banks are beginning to work proactively with homeowners in trouble to stave off foreclosure.
Bottom line: Wheaton predicts an eventual 20 percent decline in housing prices. (That would include Boston, he said in answer to an audience question.)
This 20 percent drop in housing value represents a 40 percent drop in equity or wealth of the average homeowner, because so much American "wealth" is tied up in real estate. That would, he said, lead to a 2 to 3 percent drop in consumption because people tend to spend according to their perceived wealth.
However, we would probably grow out of of a 3 percent decline in consumption in two to three years, he said. "It's just not enough to cause a recession." But, according to Wheaton, the formulation underscores the need for new thinking--that lower housing prices might cause a recession, rather than a recession lowering housing prices.
Of course, many in the audience had just one question: Should I buy now?
"I would tend to wait; I don't think the bottom has been reached," Wheaton said. "When the bottom hits, it's not going to zing back. You're not going to miss the bottom."
A version of this article appeared in MIT Tech Talk on February 14, 2007 (download PDF).
