New Arizona Home Price Index aims for more accurate measure of housing appreciation

The most popular indices currently in use today, such as those developed by the National Association of Realtors, measure the median home prices nationally and in a given region. But is median price the best way to measure the trend in home prices? W. P. Carey real estate professor Karl Guntermann and research associate Alex Horenstein have developed a new index for Arizona — this one based on repeat sales.

Looking for a more reliable way to measure the trend line in home prices, researchers at the W. P. Carey School of Business have developed a new single-family housing price index based on repeat sales. The most popular indices currently in use today, such as those developed by the National Association of Realtors, measure the median home prices nationally and in a given region.

For example, in September, the Washington, D.C.-based National Association of Realtors reported that for the prior month, the national median existing-home price for all housing types hit $224,500, up 0.2 percent from the same month in 2006.

Despite the ubiquity of these numbers in the national press, real estate professor Karl L. Guntermann suggests median home prices are not the most reliable of measures, mostly because house "quality" is not constant. "Real estate is heterogeneous. If you are trying to measure pure price change, you have a problem dealing with that heterogeneity," he says. Guntermann and research associate Alex Horenstein developed the new index together.

Skewed data

The best illustration of heterogeneity is that the trend has been to build larger homes. If you compare median home prices of 10 or 15 years ago to today, the sales of the larger homes skew the data. Crocker Liu, director of the Center for Real Estate Theory and Practice at the W. P. Carey School of Business, uses a more obvious example:

"If in one month, only expensive homes in Scottsdale were sold, but in the next month, only the cheap homes were sold, the median price would be reported as having declined, but that would only be because different houses were selling. It would mean nothing at all about true home price appreciation." Somewhat picturesquely, he adds, "the problem with median home prices is like having your head in the freezer and your toes in the oven. You get an average body temperature, but it really doesn't tell the whole story."

Introducing the ASU-RSI

The use of repeat sales for the same house is considered the most reliable way to estimate price changes in the housing market because the house "quality" issue remains constant. In other words, since repeat sales compare the price of a single house against itself at a varied points in time, the numbers don't incorporate different homes with different "quality" factors.

In the retail industry, analysts look at same-store sales to judge whether a particular chain is doing better or worse. Using repeat sales of homes is the same concept, says Liu. "The only true test of the market is repeat sales." The data for the new repeat sales-based index, which is being called the ASU-RSI, comes from Mesa, Arizona-based Ion Data/Data Express, and mirrors that of the S&P/Case-Shiller index.

New home sales are excluded because new home buyers often indulge themselves with a lot of added features — a new yard, expanded garage, new shutters, etc. — and if the home is sold just a few years later, the higher price is more attributable to improvements than to appreciation. Following S&P/Case-Shiller, the cleaning process used with the ASU-RSI excludes pairs where the first sale involved new construction and pairs with sales within six months of each other.

With the ASU-RSI, transaction sale prices less than $5,000 are dropped and pairs with more than 60 percent annual changes are also excluded. The house price data used in the S&P/Case-Shiller index starts in January 1989. If one looks at the ASU Metro Phoenix Repeat Sales Index (RSI) from January 1990 to June 2007, the data points on the graph (see Links below) show change for a particular month as compared to the same month in the previous year.

The beginning of the graph corresponds with the great real estate recession of the late 1980s-early 1990s, including the savings and loan crisis, the advent of the Resolution Trust Corp. and recession. From January 1990 through January 1992, the repeat sales index was almost always in the negative range: sale prices for individual homes were less than the same month the prior year. The index doesn't again hit negative territory until March 2007.

The most striking feature of the Metro Phoenix RSI is the steep cone shape on the right side of the graph showing the sharp appreciation in home prices over the past few years. The cone shape indicates that beginning about July 2002 percentage change on the sale price of the same home as compared to the same month the year before began accelerating from less than 10 percent to over 40 percent. After hitting a peak in acceleration around September 2005, prices were still climbing but the percentage increases were on a rapid decline.

Drilling down into the Phoenix metro area

S&P/Case-Shiller does a repeat sales index for Phoenix, but ASU-RSI massages the numbers more profoundly and can redefine the data for individual Phoenix metro areas, ranging from regions to individual cities. This drilling down of data showcases emerging contradictions in the overall market.

Scrutinizing the Regional RSI graph, which looks at corresponding trend lines for Phoenix's Central, Northeast, Southeast, Northwest and Southwest regions, one sees that the Northeast sector (including the tony submarkets of Scottsdale, Paradise Valley, Cave Creek, Carefree and Fountain Hills) exhibited surprisingly lower rates of appreciation than all other areas of the Phoenix metro.

However, when all regional sectors entered negative appreciation territory in March 2007, there was one exception — the Northeast sector — which has floated steady, neither losing nor accelerating percentage price changes. "This could indicate the Northeast metro has stabilized," says Guntermann. The same graph shows the Southeast region (Mesa, Tempe, Chandler, Gilbert) exhibited the highest percentage changes, hitting almost 50 percent around September 2005.

As can be expected, the slowing of the percentage change appreciation was quickest here and this region entered negative territory ahead of all others, at the end of 2006. This year, the Southeast region's rate of negative percentage change is one of the deepest (along with the Northwest and Southwest) of all the Phoenix metro regions.

Says Guntermann, "the regions that had the highest rates of appreciation are having the strongest rates of declines." He adds: "The trend line for metro Phoenix is downward and accelerating. Considering how rapidly prices increased the last couple of years, it could be that this decline will last a fairly long time. The good news is that the rate of decline on an annual basis is still in the single-digit range."

The W. P. Carey School is home to the Center for Real Estate Theory and Practice, which sponsored the development of the ASU-RSI Index. The center will host a national real estate conference on Feb. 22, 2008 in Phoenix entitled "Reading the Tea Leaves: Remembering the Past, Facing the Future."

Bottom Line:

• The most popular indices currently in use today measure the median home prices nationally and in a given region. The ASU-RSI one based on repeat sales.
• Median price does not take into account the heterogeneity of real estate, the quality differences between homes for example, or trends such as the increasing size of single family residences.
• Tracking repeat sales to gauge the real estate market is analogous to the same-store sales that are used to analyze the retail sector.
• The most striking feature of the Metro Phoenix RSI is the steep cone shape on the right side of the graph showing the sharp appreciation in home prices that began in July 2002 and peaked in acceleration around September 2005.