Two Days Apart, Two Reports Tell Two Stories About Prices

By Kenneth R. Harney
Saturday, September 8, 2007

How worried should homeowners or sellers be? Looking at two nationally quoted measures of house values, you might be perplexed.

At the end of August, Standard & Poor's Case-Shiller national home price index reported that prices fell by 3.2 percent from the second quarter of 2006 to the corresponding period this year. Declines in property values in some metropolitan areas were much more severe -- 11 percent in Detroit, 7.7 percent in Tampa, 7.3 percent in San Diego, 7 percent in Washington, 4 percent in San Francisco and 3.7 percent in Boston.

The sharp drop in the Case-Shiller index grabbed headlines, helped rattle Wall Street and fed fears that housing values might be unraveling in many of the country's biggest markets.

Two days after the Case-Shiller announcement, however, the Office of Federal Housing Enterprise Oversight, a federal agency that conducts home pricing surveys in more than 300 metropolitan areas, came out with its own quarterly report, and numbers startlingly different from S&P's. Rather than a 3.2 percent year-to-year decline, OFHEO's index found a 3.2 percent average gain in prices nationally. Although the agency reported price depreciation in 61 markets, prices were up in 226 others.

On a market-by-market comparison, OFHEO's numbers rarely agreed with the Case-Shiller findings: Prices in metropolitan Washington were up by 1.2 percent for the year, not down 7 percent. Chicago prices were up by an average 3.7 percent, not down by 0.7 percent. Los Angeles home prices gained by 2.1 percent, rather than dropping 4 percent. Miami homes didn't lose 4.8 percent but gained 7.5 percent.

In some parts of the country, the government's index moved in the same direction as Case-Shiller's, but by different magnitudes. Rather than San Diego's 7.3 percent decline as reported by the Case-Shiller index, OFHEO found a 3.3 percent drop. In San Francisco, rather than a 4 percent decline, OFHEO reported less than a 1 percent drop. Both indexes also reported price gains in Seattle, Dallas, Charlotte, Atlanta, and Portland, Ore.

So which has it right? How could major statistical surveys, covering hundreds of thousands of home transactions, come to such divergent conclusions?

Take a closer look: It's all about the underlying data and the purposes of the two indexes. The government's index taps into a vast housing database -- the combined new-loan and refinance transaction records of mortgage giants Fannie Mae and Freddie Mac -- but omits transactions on upper-bracket houses and jumbo loans, above $417,000.

It also omits government-insured loans (by the Federal Housing Administration and the Department of Veterans Affairs) and has only limited coverage of houses bought by people with subprime credit and home mortgages using various "exotic" terms that spurred sales in dozens of areas during the boom. OFHEO's data also include refinancings, which often have more generous appraisals than do actual home sales, but exclude condominiums.

The Case-Shiller index includes no refinancings and covers the full gamut of loan types, including the exotics, limited-documentation and subprime loans that Wall Street bought and packaged into securities. In that sense, the S&P database ranges more broadly across the home-loan universe.

However, the Case-Shiller index -- devised to facilitate creation of a trading marketplace for housing futures and financial hedging vehicles -- has its own limitations. For example, it contains no data from 13 states and has incomplete information from 29 states because of legal restrictions and lack of electronic records from local courthouses.

OFHEO chief economist Patrick Lawler considers that a criticaldifference. By his calculation, the 13 states missing from the Case-Shiller numbers "averaged 6.3 percent gains over the past four quarters -- these are now some of the strongest housing markets," he said, but they are nowhere reflected in the 3.2 percent national depreciation rate reported by S&P.

Lawler also notes that the Case-Shiller numbers are weighted -- a $700,000 house counts twice as much toward the index as a $350,000 house. The OFHEO numbers are not weighted.

Terry Loebs, managing director of MacroMarkets, said the missing states do not greatly affect the Case-Shiller index. For example, Wyoming, which OFHEO ranks as the second-fastest appreciating state with an average annual price gain around 13 percent, represents "just one-half of 1 percent" of the national housing value pool, Loebs said. Not having data from Wyoming "is not a significant drawback." Loebs also said his company's index "represents over 70 percent of the value of houses" nationwide, far beyond OFHEO's scope.

Bottom line: Don't overreact when you see big drops -- or jumps -- in these indexes. They are measuring different things, and no national index gets down to the nitty-gritty: what's happening to property values in your Zip code, micromarket or neighborhood.

Kenneth R. Harney's e-mail address is

© 2007 The Washington Post Company