patrick.net

Here's an interesting comparison of the Shiller 100-year index and the CoreLogic HPI from calculatedriskblog:

The upward slope of Real House Prices
http://www.calculatedriskblog.com/2011/05/upward-slope-of-real-house-prices.html

McBride's opinion is that:

"In many areas - if the population is increasing - house prices increase slightly faster than inflation over time, so there is an upward slope for real prices."[...]

I think they are both showing that prices are not far above the historical lows. Prices might overshoot to the downside because of supply and demand issues; there is a large overhang of vacant housing units and many distressed properties still coming on the market, plus demand is soft with weak employment, fairly tight financing, negative home buying sentiment and some usual buyers excluded because of credit issues. [...]

BUT...

usually near the end of a housing bust - after nominal prices stop falling - real prices decline slowly for a couple more years, and we will probably see that this time too. Of course, right now, nominal prices are still falling.

His issue is that Shiller parameterized his model based on FHFA data from the 1975-1985, which McBride believes did not reflect the true state of the housing market at that time. I can't say which model is better (Shiller = real prices flat or CoreLogic = real prices increasing).

I have a few of issues with this analysis though:

1. I'd like to see how the line slopes were calculated/modeled for those very curvy curves

2. I'd like to see how these indices compare on smaller scales (like San Jose versus Phoenix), relative to census population data and current population trends

and

3. I'd like a better explanation of where CoreLogic's data came from over the contested period.

#housing