The median is one of several market indicators and as such is a perfectly legitimate indicator of the market. How can one use it to justify the fact that the market is overvalued and due for a correction up to mid-2005, then toss it out as irrelevant in 2006 ?
Having said that, it is not accurate as a measure of what a particular home would sell for, nor does it adequately capture the trajectory of the market. It does not tell you what your house is worth. But as an indicator of market conditions ? Sure, it’s reasonable.
The median price has been used to demonstrate the following :
1. There was a large run-up in SD real estate above fundamentals (rental rates and incomes) from ~2000 to 2005.
2. There was a softening in sale prices starting in 2005.
3. That softening has led to declines in prices and continues to persist through November 2006.
Combined with sales activity, I find the median to be one of many useful barometers of the market, particularly for SFRs in established areas. If we toss out all of the useful estimators that have flaws, then we’ll be left with only anecdotal evidence (which is also quite useful, but not without other confirmatory evidence).
Personally I found the use of median prices by Rich in the Primer on this site to be quite useful for illustrating the extent of the residential real estate bubble in San Diego. It may not be perfect, but if it’s a legitimate parameter for making the case of the bubble, it ought to be useful for evaluating the extent of the bust.
