Recent data suggest that the housing market topped out last summer and that home sales and price increases have slowed. Although the picture is confused by an unusually warm January and by rebuilding after Katrina, which together created a spike in housing starts, the overall picture is becoming clear as the spring data become available.
So far, it appears that the slowdown will be gradual, as the Federal Reserve hopes. The risk of a sharp break remains, however, especially if mortgage rates spike.
There are two dangers to the economy from the end of the housing boom: First, we expect housing starts to drop from 2.07 million to 1.91 million this year, despite about 100,000 new housing units expected as replacement for hurricane destruction.
Second, higher interest rates and lack of further appreciation will restrict homeowners' ability to take cash out of their houses through cash-out refinancing or home-equity loans, thus hurting spending power. If the declines are gradual, the slowdown will be moderate, as the Fed wants.
HOME PRICES SOFTEN. Home sales have weakened slightly over the past few months, abstracting from the weather-related bounces. More telling, the number of unsold homes on the market has surged, with the supply of unsold existing homes at 6.5 times monthly sales in May, up from 4 times monthly sales a year and a half ago. Although the inventory of unsold homes remains near its historic norm of six months' worth, the rise shows that the market is slowing sharply after its incredible five-year run.
The annual rate of 6.67 million existing homes sold in May was down 6.6% from a year earlier, yet still near the 2005 record annual sales of 6.78 million. However, the fact that sales are down year-over-year for six consecutive months does seem to indicate the start of a more negative trend.
Sales prices continue to rise (up 6% from a year earlier in May), but this reflects in part a change in regional mix. In any event, the rise in the inventory of unsold homes suggests houses aren't moving quickly enough at these prices.
The picture for new homes is similar, but weather has been more of a factor. Housing starts surged in January to an annual rate of 2,307,000, but that reflects weather, not economics. Starts retreated to 1,957,000 in May, down 3.8% from a year earlier and below the 2005 total of 2,068,300. New home sales dropped to an annual rate of 1,234,000 in May, down 5.9% from a year earlier, while the inventory of unsold new homes rose to a six months' supply, vs. 4.2 months a year earlier.
HOUSING ACTIVITY OUTLOOK. Housing remains affordable by historical standards, however. In mid-2004, low rates had improved affordability to the best level since 1973, but higher home prices and rising mortgage rates have pushed it back to average.
Starts will also be supported by the need to rebuild homes lost in last year's hurricanes. About 350,000 homes were destroyed; we expect fewer than half to be rebuilt this year, but that may be 8% of 2006 housing starts, a significant boost.
The net result is likely to be a moderate drop in housing starts, to 1.91 million in 2006 and 1.73 million in 2007. The 23-year high was 2.07 million, reached in 2005. In most economic cycles, housing starts drop under 1 million at the trough, although 2001 was a major exception, when the Fed's prompt action in cutting rates held the low at 1.5 million.
THE ECONOMIC IMPACT. The weaker housing market will affect the economy in two ways: First, falling construction activity will have a direct negative impact. Second, higher mortgage rates and an end to strong home appreciation will restrict Americans' ability to take cash out of their homes.
Construction activity is expected to fall only 0.1% (corrected for inflation) this year, but 8.6% next year. The 2006 data are helped by the strong start to the year, the carryover of activity, and the hurricane rebuilding. In 2007, these supports disappear. But the direct impact on the economy will be modest, since residential construction accounted for only 6.1% of gross domestic product last year, and represented 0.4 percentage points of last year's 3.5% real GDP growth.
This year, the contribution will be negligible (a drop of less than 0.1 percentage point), while next year's contribution will be –0.6 percentage point. The housing sector will thus cause a swing of a full percentage point in real GDP growth between 2005 and 2007—the total slowdown expected for the economy (3.5% to 2.4%).
The second impact comes through the reduced ability of Americans to use their homes as giant ATM machines. Homeowners have taken nearly $700 billion in home equity over the last two years in cash-out refinancing or home equity loans, or about 3.5% of disposable income. Based on survey data from the Mortgage Bankers' Assn., we estimate that one-third of that was spent on home renovation and is thus included in the impact of residential construction discussed above.
The second most prevalent use of this money was education; we expect that education funding not raised from home equity loans will instead be raised in student loans.
There is still ample home equity to borrow against. The overall loan-to-value ratio in the housing market has remained near 43%. The ratio of mortgage debt to disposable income has risen to a record 98%, but that reflects the rise in home ownership and prices, not an increase in leverage.
Households still have room to raise cash from their homes, but it will cost more than in the recent past, and that should discourage borrowing. The weaker housing market and rising mortgage rates will thus deter consumer borrowing and spending, but probably not as much as some people believe.