Los Altos Real Estate Blog

A significant number of news articles quote the Case-Schiller Index as a sign of how good or bad the overall national market is doing.  At times when the index is up, people are happy, when it's down they're unhappy.  The problems with the Case-Shiller Index include not covering sub-markets, not using price bands that really represent the market, and using "sales pairs" which only show a statistically small portion of the market.  Their definition of "high-end" in each market is a joke, they don't break down sub-markets of the markets they track, they exclude markets that are doing well, and they don't give a reasonable timeframe for reporting.  It takes them two months to publish their report.  Really!?  Two months ago might as well be forever.  Markets change and all the Case-Shiller index does is give people headaches and stomach acid. I used to respect their index until they showed the SF Bay Area as having problems.  One reason is that they don't include Santa Clara County in their calculations.  Hello!  We have some of the best performing markets in the country.  A significant portion of the local market is up...WAY UP and the Case-Shiller index is missing it.

So, knowing the flaws in Case-Shiller and realizing that it is lacking in so many ways, I'm going to create a new index.  Since using last names seems to be popular, I'll call it the Robertson Home Value Index (RHVI).  If anyone has better ideas, I'm all ears because the name doesn't matter, it's all about reporting statistics that make sense.  Here's how my index will work so that we can see an accurate accounting of how well the market is doing.  Here are the elements of my index:

Price Bands:  Prices are relative, everywhere.  An entry level condo in San Jose can be priced the same as a 4-bedroom home in Ohio.  It's not fair to compare the two.  So, the RHVI will use the MEDIAN SELLING PRICE of the local market as a baseline.  It will then use that baseline to set high and low bands.  There will be 4 bands - high-end (200+% of the median), upper mid-range (125-199% of median), mid-range (75-125% of median), and low-end (74% or less of median).  This gives my index 4 relative bands versus Case-Shiller's 3 fixed bands and shows a true "high-end" of the market.

Co-Located Sub-Markets:  Or in easier language, local markets that are peripheral to the primary market.  Using the Bay Area as an example, the RHVI will use San Francisco as a primary market and publish stats only for that market.  It will then include the top 3 local neighboring markets based on their relationship to the SF Market.  Which local markets are included will be done with the same relative calculations as those used to define the price bands noted above.  That way, all the calcuations will be relative.

No Sales Pairs:  What the economists forget in their sales pair methodology is that a home selling twice only shows a subset of the market behavior.  A home purchased for a high price in 2008 that sells today at a good market price might look like it has dropped in value.  Which is true, but it misses out that it sold in line with other comps.  Sales pairs fail to show a general market trend.  By using all sales, true trends can be shown.

Obviously this could take a long time to explain but I'm working on a spreadsheet using all the sales data in the area.  I'll see what I can come up with but in my estimate, whatever the final product, it will show a more accurate picture than Case-Shiller.