Is the S&P/Case-Shiller index flawed?

You might want to get prepared for a battle between Radar Logic's Jonathan Miller and Dr. Robert Shiller, co-founder of the S&P/Case-Shiller index. No, it's not the kind of fight where we'll see anyone take a swing, but there could be some damaged pencils and an increase in carpal-tunnel syndrome among staffers at both companies attempting to defend the words of their leaders. From a Reuters story:

Standard & Poor's/Case-Shiller indexes do not adequately reflect U.S. home prices, mainly because the monthly reports are based on repeat sales and exclude new development, said Jonathan Miller, executive vice president and director of research for Radar Logic.

"The basic premise is flawed," Miller, speaking at the Reuters Housing Summit, said of the Case-Shiller indexes.

S&P-Case/Shiller produces a closely watched monthly index as well as a quarterly gauge of home prices in the top 10 and top 20 metro areas.

Using a repeat sales method, the home price change of a house sold now would compare with the last time it was sold, which could be years ago, Miller said.

"When you do a repeat sales index, you exclude all new development because there was no prior sale," Miller added. "They also exclude condominiums, so you are covering the New York City market, and you're not including condominiums, yet you're representing that that's the New York number."

The biggest problem with excluding new development, he said, is that it influences prices of existing properties in that market.

The Case-Shiller indices are a "great concept if you are measuring existing housing sales in suburbia," Miller said. "It's a very academic approach, and it's technically correct. However, it doesn't represent the actual cities that are being covered accurately, in our opinion."

S&P is not taking these criticisms lightly:

David Blitzer, managing director and chairman of the Index Committee at Standard & Poor's, countered that the repeat sales method captures the broadest swath of the housing market and provides a uniform source of comparison.

"To combine new homes and existing homes in a single index is a mistake because it's a different phenomenon, a different kind of animal," he told Reuters in a telephone interview.

"The existing homes are the bigger and the more important segment" of the housing market, he added.

Sales of existing homes roughly outnumber sales of new houses by at least five to one, Blitzer said.

And that is true -- on a national scale. But what about specific markets like California's Inland Empire where the ratio of new-to-existing home sales is higher? In the case of S&P/Case-Shiller it doesn't matter because they don't track the IE separately.

So how is Radar Logic different?

Radar Logic is a data and analytics business based in New York that produces a daily "spot" price for residential real estate in major U.S. metropolitan areas.

It tracks housing markets by publishing daily the price per square foot in 25 metro areas, Miller said. That compares with the Case-Shiller indexes, which are publishing once a month using a 90-day moving average, he added.

"The reason these indexes are being created," Miller said "is to enable trading of financial instruments on Wall Street. You want to reflect subtle changes in markets, and it's very difficult to do that if you're doing a 90-day moving average and giving 12 data points a year."

Another shortfall is that Case-Shiller "doesn't factor in remodeling, expansion of footprint, which has been a very significant trend in housing, especially over the housing boom," Miller said.

Expect more on this "battle of the indices" to come.