Case-Shiller: Major Bleeding Continues for Low Tier Home Prices

It’s time for our monthly check-in of the S&P/Case-Shiller Home Price Indices (HPI). The Case-Shiller data is generally considered to be the most reliable measure of overall home price changes for a region, since they only consider repeat sales of homes when calculating their index, instead of looking at all the homes that sold in a given month.

For the full source data behind this post, hit the S&P/Case-Shiller website. For a more detailed explanation of how the Case-Shiller Home Price Index is calculated, check out their methodology pdf. Also remember that the data released on the last Tuesday of a given month is for the period two months prior (i.e. – October data is released in December).

Here are the basic Case-Shiller stats for the Atlanta area* as of October:

October 2010
Month to Month: Down 4.2%
Year to Year: Down 6.2%
Prices at this level in: April 2000
Peak month: July 2007
Change from Peak: Down 24.3% in 39 months
Low Tier: Under $129,786
Mid Tier: $129,786 to $228,473
Hi Tier: Over $228,473

All twenty metro areas tracked by Case-Shiller saw a decrease in their HPI between September and October (vs. 18 August to September). Only four markets posted year-over-year gains: San Francisco, San Diego, Los Angeles, and Washington DC.

Here’s a look at the latest local tiered data, back through 2000:


And here’s a closer look at the recent changes, with the vertical and horizontal axes zoomed in to show just the last year:


The low tier in Atlanta continued to lose ground at a ridiculously fast rate in October. Month to month, the low tier was down an unbelievable 10.0%, the middle tier fell 1.8%, and the high tier decreased 2.7%.

The drop in the low tier was so severe I was literally having a hard time believing that it’s legit. Therefore, I contacted a representative at Case-Shiller, who put me in touch with David Stiff, their Chief Economist and Vice President of Quantitative Research. Here’s what he had to say after looking into the numbers:

I have checked the low tier Atlanta S&P/Case-Shiller index – there are no errors in the calculation of the index.

…sale pair counts have declined in Atlanta. It is possible that this is due to delays in deed recording. If that is the case, then it is possible that there will be upward revisions to the Atlanta low tier index in future updates.

However, another explanation for the decline in sale pair counts and the Atlanta low tier price index is the expiration of first-time homebuyer tax credit. The decline in the S&P/Case-Shiller index is consistent with other data for Atlanta – the NAR price median (for all single-family homes) dropped by 7.5% between 2010:Q2 to 2010:Q3 and sales activity for Georgia dropped by 14%.

The low tier indexes will be more volatile than the middle, high, and aggregate price indexes. During the housing bubble, property flipping and fraudulent mortgage lending activity tended to be concentrated in the lower-priced segments of metro housing markets. As a consequence, larger proportions of low-priced homes have been foreclosed upon, resulting in larger price declines in this segment. During the home price correction, investor purchases of bank-owned properties and the stop-start nature of government programs designed to bolster housing markets (e.g., HAMP, tax-credits) have added more volatility to the lower-priced segments.

So, despite the incredible rate of decline, the data does appear to be legit.

Here’s a chart of Case-Shiller HPIs for all the markets that Redfin serves:


Here’s our peak decline chart, in which we line up the peak Case-Shiller HPI value for each of Redfin’s markets, so we can see how long each market has been declining, and how much it has dropped from the peak.


Here’s the flip side of the peak decline chart—a graph since January 2009, indexed to January 2009 = 100%:


With the post tax credit home price double dip now in full effect across the board, I think it’s about time to retire this last chart.

Methodology: The Case-Shiller index tracks price changes in sets of homes of similar size and style to better determine changes in what people are willing to pay for the same home over time. If data is available from an earlier transaction for the same home, the two sales are paired and treated as a “repeat sale.” Repeat sales that are too far apart, sales between family members, lot splits, remodels, and property type changes (e.g. from single-family to condos) are excluded from the calculations. All remaining repeat sales are totaled together and weighted based on the time between each sale, then the data for the most recent three months is averaged together to create a given month’s index value (i.e. – September’s index represents the average of the data from July through September).

The three price tiers plotted in the charts below simply represent the top, middle, and bottom third of all sales, based on the initial sale price. In other words, if there were 3,000 sales in the three-month period, 1,000 of them would be in the low tier, 1,000 in the middle tier, and 1,000 in the high tier, by definition.

*[Case-Shiller defines Atlanta as the Atlanta-Sandy Springs-Marietta, GA Metropolitan Statistical Area, which includes all of the following counties: Barrow, Bartow, Butts, Carroll, Cherokee, Clayton, Cobb, Coweta, Dawson, De Kalb, Douglas, Fayette, Forsyth, Fulton, Gwinnett, Haralson, Heard, Henry, Jasper, Lamar GA, Meriwether, Newton, Paulding, Pickens, Pike, Rockdale, Spalding, and Walton.]