AMID SUCH RAMPANT SPECULATION that house price bubbles have formed and are on the verge of collapse, builders are losing sleep. After all, falling prices would erode the hefty margins that home builders in rapidly appreciating markets have come to enjoy. And falling prices would slow sales as people that rushed to buy when prices were surging retreated out of fear of continuing declines.

But the house price situation is not nearly as dire as one might conclude from reading the headlines. While some markets are at risk of a correction at some point, the conditions that might cause significant declines in most markets aren't in sight. In fact, in 90 of the largest 153 metro areas, house prices have grown in line with or no more than 30 percent faster than per capita income growth over the past 20 years (see “Metro Metrics,” page 26).

Growth ahead of income has been spurred since 2000 in part by low nominal interest rates. Decreases in these rates have historically helped drive prices up but have not figured prominently in large price corrections.

Still, the pace of recent price gains in some markets is unsustainable. At the rate that prices went up on average last year in Las Vegas, for example, house values there would double every three years if sustained. In Los Angeles, San Diego, and Reno, they would double about every four years. Price appreciation in these areas must slow soon to avert future declines.

But it takes a crystal ball to forecast where, when, and by how much house prices will decline. It turns out our models are pretty good at predicting future house prices in many states and metropolitan areas. House prices move nearly in step with per capita income growth in 39 of the 50 states, and most metro areas. It is because of this that in most of middle-America, house price growth historically has not been that dramatic and depreciation even less so.

PROBLEMS PREDICTING Unfortunately, our ability to gaze into the future confidently with just one variable falls apart in some of our most populous metropolitan areas where a significant amount of home building takes place. The predictive power and reliability of even highly sophisticated models with slews of variables is pretty poor for metropolitan areas in Florida, California, Hawaii, and along the coasts of Oregon, Washington, and the states that stretch from Massachusetts to Virginia. These are also the places that tend to have the most pronounced housing cycles—large upswings are followed by years of slow or slightly negative growth.

What causes more volatile cycles in these places is not entirely clear, but compelling evidence is mounting that land and development regulations and restrictions play an important role. Housing supply cannot smoothly adjust to increases in demand during an upturn.

METRO METRICS As a result, buyers bid up prices as they chase a limited supply. At some point, risk-takers seeking to exploit the soaring prices pile into the market, buying homes on speculation in hopes of a quick gain. When and if an economic recession of real magnitude comes, then demand evaporates, supply comes on line after a long lag, and prices deflate. But it also tends to drive up prices and wages higher, on average, over the long run. Long-term owners benefit from these trends even if some can get whipsawed by temporary house price deflation.

SEEING CYCLICALITY House prices do not fall in a vacuum. Something has to bring them down. The two principal culprits are job loss and overbuilding. Both lead to a glut of homes for sale that causes sellers to accept lower prices. Furthermore, house prices behave to some degree like that old Jimmy Cliff song: the harder they come, the harder they fall. When house prices skyrocket, they have that much farther to fall to get back in line with long-term fundamentals. Some combination of job loss and overbuilding after a period of strong economic growth and escalating home prices, for example, brought down house prices in places such as Boston, Houston, Los Angeles, and New York.

So where are we in the cycle? There is hard evidence emerging that speculative buying is adding fire to some already overheated markets. Nationally, Freddie Mac reports that the number of homes the enterprise financed that were flipped within a year of purchase inched up from 5 percent to 6 percent from 1998 to 2003. In that same five-year span, the share flipped within one and two years climbed from 11 percent to 14 percent. And though the economy is adding jobs, it's at a disappointing pace, with some places posting small losses. It's simple to see that house prices are ahead of income growth in several places—house price-to-income ratios in 34 metro areas are at their highest level on record.

Learn more about markets featured in this article: Los Angeles, CA.