I would be interested to see the historical data that supports the conclusion that “the percentage tends to stay the same”.
Here’s what I find out from Zillow and their graph per zip code. I’m using Mira Mesa and Carmel valley as an example. In 1997, the median was $144k in MM and $239k in CV. That’s a spread of 66%. in 2006, at the peak, MM was @ $537k and CV was at $906k. That’s a spread of 68%. So at least comparing to MM to CV, the percentage stays pretty constant.
Comparing between PQ and MM, the # is further apart. in 1997, the gap is 39% while now, in 2006 it’s @ 24%. Comparing 92126 vs 92128, it’s 22% in 1997 and it shrinks to 7% in 2006. 92126 vs 92122, it’s 25% in 1997 and 15% in 2006. 92126 VS 92037, it’s 115% in 1997 and 99% in 2006. 92126 vs 92127, it’s 34% in 1997 and 31% in 2006. 92126 vs 92131, it’s 50% in 1997 and 33% in 2006. So, as you can see, some area held their premium pretty constant while other lost a premium anywhere between 10-17%. The reason why I use Zillow graph is that their algorithm is the same for all area, so it removes one variable. They also uses the same data in their calculation as well. So it’s as close as you can get to these kind of info. So, yes, when we hit this bottom, area like 92131 might have their 50% premium again compare to 92126 instead of their 32% premium now. But I highly doubt they spike to 100+% because they don’t fall while other less desirable area fall more. At some point, people will ask, is it really worth paying 2X to live in Scripps Ranch compare to Mira Mesa. That’s my 2.5 cents.