Use access key #2 to skip to page content.

EScroogeJr (< 20)

Case-Schiller: a really useless index



May 27, 2008 – Comments (6)

The Case-Schiller index  is often quated as an evidence that prices are tumbling. Now, this conclusion is dead wrong. There is almost zero correlation between the index and your chances of getting a reasonable price for a house. Why? For two reasons. First, the Case-Schiller is a weighted index. In other words, when a $10M castle gets 1% cheaper, and 100 $100K houses get 1% more expensive, the index thinks that prices are stable. So when the  top decile that did get a little frothy corrects, it will produce a dramatic change in the index without necessarily changing things on the ground.  The second reason is that the Case-Schiller  tries to be a hedonics index that measures quality changes, such as footage. The footage, of course, always grows in time, meaning that C-S countinuously makes housing appear cheaper than it is. This fotage increase, however, cannot be attributed to changing consumer preferences, becuase it's driven almost exclusively by the building departments that prevent developers from listening to the market where the demand is for smaller houses, forcing them to build in the luxury segment only. Now that some of these guargantuan houses are on sale, the index naturally reflects this. However, there is little consolation in this for 90% of buyers, because a discounted McMansion is still a McMansion and still way out of the affordable range. While C-S may be a usefull gauge for the new construction market, it is quite useless for the average buyer. As a bone to the bears, there's no use denying that C-S is down substantially.


"NEW YORK (AP) -- U.S. home prices dropped at the sharpest rate in two decades during the first quarter, a closely watched index showed Tuesday, a somber indication that the housing slump continues to deepen. Standard & Poor's/Case-Shiller said its national home price index fell 14.1 percent in the first quarter compared with a year earlier, the lowest since its inception in 1988. The quarterly index covers all nine U.S. Census divisions. Prices nationwide are at levels not seen since the third quarter of 2004, according to Maureen Maitland, a S&P vice president. However, the index is still up 60 percent versus 2000." 

6 Comments – Post Your Own

#1) On May 27, 2008 at 11:46 AM, goldminingXpert (28.83) wrote:

Blah blah blah... got a load of homebuilding stock you're trying to unload?

Report this comment
#2) On May 27, 2008 at 11:56 AM, EScroogeJr (< 20) wrote:

goldminingXpert, you don't have comment on Caps, you know. There are Yahoo message boards for the likes of you.

Report this comment
#3) On May 27, 2008 at 12:01 PM, goldminingXpert (28.83) wrote:

I just don't understand why people are delusionally trying to remain optimistic about homebuilders and the real estate market. Doesn't a chart of XHB tell the true story clearly enough? I've made a ton of CAPS points shorting homebuilders, and as long as stupid SPF-style deals continue, I shall gain more CAPS points this way. The real estate market is 1991-Japan bad, yet people think it is only 1989-Texas bad or not even that.

Report this comment
#4) On May 27, 2008 at 12:03 PM, goldminingXpert (28.83) wrote:

EscroogeJr: Residential construction: 12 active picks. 0 percent accuracy!!! -217 points!!! You are not worthy of blogging on said subject. Sorry.

Report this comment
#5) On May 27, 2008 at 11:06 PM, eskatonic (28.60) wrote:

I find it insightful when anyone intelligently takes apart an index.  most people blindly follow an index without really understanding what it really is or how it really works.  they just have some vague notion, like CS is supposed to tell you if housing is good or bad.


way to go ESJ!

Report this comment
#6) On June 02, 2008 at 9:52 AM, saunafool (< 20) wrote:


I really wonder about you. You are one of my favorites because you speak your mind and have a wholly independent view. I think you're wrong about the market, but CAPS is great because time will tell. (I also wonder if perhaps you are deliberately taking the side of the trade no one wants because CAPS is more fun for you that way.)

Anyway, assuming you really don't like the CS index, I personally think its the best index for a market where the product constantly changes. If you look at what the NAR puts out, they will say the price of a home doubles every 10 years. Of course, they look at new home prices and they have to stop their data series in 2007.

One of the things they do not account for in their "doubling every 10 years" is that new houses have been constantly getting bigger.

Looking at the general government housing price index, they just look at average selling prices in a given region and assume the product mix is not changing--leading to exactly the type of data skewing you complain about with the CS index. In other words, if the high end of a market stays healthy, but the entry level market implodes, the government index will say prices are rising when in fact the market is falling apart. (This is exactly what happened in CA from the end of 2006 through 2007--the numbers showed prices going up while tons of entry level inventory piled up unsold.)

The CS index does the best it can to address the weaknesses of overall sales numbers by looking at repeat sales of the same house and accounting for upgrades to the house. Let's say you buy a 1000 square foot fixer upper in Silicon Valley for $600,000 (and for that price, it is going to be practically a tear-down) and sell it 3 years later for $1.2 million. After you buy it, you spend $400,000 renovating it and expanding the size to 1800 square feet. What is the appreciation of the house?

According to the NAR, the value of the house doubled.

According the the CS index, the value increased by 20%.

If I'm making decisions with my money, the CS index looks a lot better to me.

Obviously, this is overly simplified, and I don't mean to imply that you think the NAR or government numbers are any better. Just from what I know about the CS index, Shiller was trying to account for factors that were not included in typical housing indexes, such as products changing and people putting huge amounts of money into the houses to improve the value.  On that score, the CS index seems to be more accurate than anyting else around.

Report this comment

Featured Broker Partners