Only One Person Knows a Home's Value: Its Buyer

House-Price Index Readings Can Be Inflated, Built on Shaky Foundations and Far From the Right Neighborhood

The Wall Street Journal Online
By CARL BIALIK
http://online.wsj.com/article/SB122722235538745845.html?mod=djempersonal

The good news is your home may be worth more than the rock-bottom price that your neighbors' houses fetched. The bad news: No one but you might think so.

The one point of widespread agreement in the real-estate industry is that there is no single accurate index of home prices. They are all over the map, cover different sets of homes and may exclude parts of the country or be unduly influenced by the mix of homes sold in a given month.

A sold sign is displayed in the yard of a house in Clarksville, Tenn., in October.
As the home market surged earlier this decade, the two leading indicators of home prices diverged. One didn't count homes sold with exotic or subprime mortgages, which fueled much of the bubble. These same properties are often the ones going on the auction block today at severe discounts, pulling the other home-price index down -- some say to unrealistic lows.

To address these discrepancies, indexes are going increasingly local. Other, less-well-known measures of home prices -- some of them available only to paying customers -- are adjusting to exclude homes sold by banks.

Sales of foreclosures and other distressed properties accounted for 35% to 40% of transactions in the third quarter, the National Association of Realtors said this week. The discount on such properties, often sold by banks that need to clear inventory quickly, can be 30% to 40% compared with similar properties sold by the resident, according to Damien Weldon, a vice president of credit-risk products and analytics at First American CoreLogic. The company's Loan Performance division is producing a new index without these discounted sales, a distinction that was "not important a few years ago, but now it's very important," Mr. Weldon says.

Behind the Home-Price Indexes
The numbers from home-price indexes are widely watched. The Federal Reserve uses them to measure the value of housing stock. Banks use them to determine whether mortgages are underwater and to estimate the value of homes they will have to sell after foreclosure.

But the indexes may be leading everyone astray. Just as respondents to election surveys are meant to stand in for the broader electorate, the homes being sold need to represent all homes. The problem is, producers of these price measures aren't sure that sale prices reflect the values of houses not on the market.

"People put all their eggs in the sales-price basket," says Andrew Leventis, a senior economist with the Federal Housing Finance Agency, which produces a home-price index.

"Whether the transaction pool is reflective of the entire housing stock -- nobody addresses that problem," adds Karl Case, professor of economics at Wellesley College and co-creator of the Case-Shiller Index, a competitor to the federal government's measure.

The Case-Shiller index includes properties that had subprime loans attached.

"That's the stuff that went down most substantially, and that's probably the stuff that went up most substantially," Prof. Case says.

The federal index, though, doesn't include such properties, instead accounting only for properties with financing from mortgage giants Fannie Mae or Freddie Mac. For that reason, many prefer Case-Shiller.

"I believe S&P Case-Shiller for the areas it covers," says Thomas Lawler, a housing economist in Leesburg, Va.

Case-Shiller has shown a steeper decline in markets with many distressed sales. The second-quarter year-over-year declines in San Francisco, Phoenix and Las Vegas ranged from 23% to 28%, according to Case-Shiller. But the federal gauge recorded declines of only 5.8%, 11% and 18%, respectively.

Not everyone thinks the Case-Shiller index is useful. Richard A. Smith, chief executive of real-estate broker Realogy Corp., says the index omits 13 states. "Case-Shiller as a broad index is inaccurate," Mr. Smith says.

David Blitzer, chairman of the index committee at Standard & Poor's, which publishes Case-Shiller, responds that "the sampling and data collection is as good as it can be."

"One's got to be wrong," Mr. Smith said of the dueling Case-Shiller and federal indexes. "Nobody will know until the book is written."

Yet there is no surefire way to know which index got closer to the truth. Each year since 2000, the Census Bureau has asked homeowners to report the value of their home, but "it doesn't necessarily jibe with assessment records or anything like that," says Jeanne Woodward, a Census Bureau statistician.

Another potential check on values is home appraisals. But Dr. Leventis said there are two possible sources of upward bias. One is that people often choose to get their homes appraised when they figure the value has risen sharply and they can convert some of that to cash with a refinancing. Another is that appraisals tend to overstate the value of homes, perhaps because homeowners seek the most-favorable assessment.

"I really don't see a benchmark" against which to check these home-price indexes, says Lawrence Yun, chief economist of the National Association of Realtors.

His group releases its own numbers, most recently showing prices declining by 9% in the third quarter compared with a year earlier. But that measure, unlike the others, doesn't take into account a home's sales record. So instead of comparing each property's sale price to its prior sale price, the realtors group compares the price of homes sold this month with that of homes sold last month -- even if the mix of homes has changed sharply. Mr. Yun defends the measure as "very simple to understand."

Most of the numbers that get headlines are based on metropolitan areas. Yet the housing-market picture can vary dramatically within the same region. Lynn, Mass., a suburb northeast of Boston, saw prices drop 10% in the second quarter compared with a year earlier, according to Wellesley's Prof. Case. Yet in the same period prices in Cambridge, just west of the city, rose 13%.

Integrated Asset Services, or IAS, sells estimates by neighborhood. "We are a lot more granular" than Case-Shiller and the federal index, Chief Executive David McCarthy said.

Within Middlesex County, which includes Cambridge, one neighborhood was up 12% compared with a year earlier in September, while two others were down 1% and 2%, respectively.

Fiserv Inc. uses the Case-Shiller local numbers to sell estimates for a single property.

The risk when going local is that data become sparse and a few anomalous sales may throw things off -- particularly in markets where most of the sales are distressed. IAS makes estimates based on as few as 50 to 75 transactions.

None of this nuance is captured in headlines about the latest home-price-index release, Prof. Case complains. Still, he is hopeful that home-price indexes will improve. "This new criticism that these indexes are showing different things is going to lead to a lot of research," he says.

[This article points out the problem of using national or state or even county index values. In our market area, the SCV Home Team does a much more accurate analysis of home value.]