The nationwide collapse of the housing market continues to bedevil the economy. The U.S. housing market peaked in the spring of 2007; since then, house prices have fallen an average of 16 percent. In the worst-hit regions (Miami/Fort Lauderdale, Orlando, Riverside, Phoenix, Las Vegas) prices are down between 45 and 59 percent. And those are metro averages that doubtless obscure neighborhood-level declines that are truly cataclysmic.
But of the 64 metro areas with population greater than a million, seven actually have average prices greater now than at the national housing peak four and a half years ago. Here they are with their net gain since the second quarter of 2007:
- Pittsburgh: 5.4%
- Buffalo: 5.0%
- Houston: 3.1%
- Austin: 2.2%
- Rochester: 1.9%
- Oklahoma City: 1.2%
- San Antonio: 0.1%
This list is split between the South Central U.S. and the Northeast. Indeed, the next two metros on the list are Fort Worth and Dallas. But then come Louisville, Raleigh, Indianapolis, and Nashville. So geography isn’t a complete explanation. Neither is the regional growth rate: Pittsburgh, Buffalo, and Rochester are growing more slowly than average but the Texas metros are growing much faster than average. So what do these metro areas have in common?
First, though, a quick word about where these price changes are coming from: the Federal Housing Finance Agency (FHFA) House Price Index. The FHFA is the federal agency that regulates Fannie Mae and Freddie Mac. The House Price Index is calculated quarterly based both on repeat sales and refinancings — so it avoids some of the problems of using the median price of houses that happen to sell in a given month to track the market. This index also has the advantage of longevity: it is available for the U.S. and many metros back to the mid to late 1970s.
The chart below dramatizes how different these metros’ price trends are from average. It shows the three best-perfoming metros compared to the U.S. The U.S. declined sharply from its mid-2007 peak. The three metros increased much less than average during the boom but continued to appreciate through 2008. After that, their gains stopped but reversed little — if at all.
Comparing pre-crash house price growth to that after the peak for all 64 metros with a population of a million or more makes the relationship clear: those metros with the largest price gains before 2007 had the largest price declines afterward, while those with small gains had small (or no) declines. In fact, a given metro’s price behavior before 2007 explains 74 percent of its behavior afterward. It may be that housing investment is inherently safer in some regions than in others. In these regions, you cannot expect to make a killing, but neither are you likely to get killed. The next step will be to identify factors that drive this difference.