The Case Shiller

Now, about Case Shiller...

I always tried to put a c in Shiller, because I didn't want the root word to be shill. (http://en.wikipedia.org/wiki/Shill) And the Case Shiller series has provided so many alarming headlines that it seems worthwhile to look at its main problem.

To begin with, it's worthwhile to look at how other statistics mislead (http://www.charlesbwarren.com/train_wreck.html). If you are looking at averages, mean or median, they sort of assume that the population remains similar between time periods. That isn't the case today.

That problem was addressed by Case-Shiller (C-S). If it's sale resale of the same house, then the population is fixed. No problem, right? Wrong.

First there is often a divergence between average price and sale resale indexes, at least in "particular circumstances" (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=978981, sorry you'll have to request the whole paper to find the divergence). As we are presently in "particular circumstances" it is worthwhile to note that sale-resale numbers held up better in Paris than average price. The reasoning was that the properties that were sold and resold were probably in better condition, and possibly that the sellers were more motivated to obtain the highest price because of their investment in the property. Note that France has high transaction costs and low property turnover compared to the USA.

The population in C-S is superficially the same. The physical property occupies the same land. There are, however, two gaping disparities.

First is terms of sale. I will use an example with which I am very familiar. The first sale was in summer 2006 for something over $600,000. No down payment; there was a large seller-carried 2nd deed of trust. Even in revenue-hungry California where price is almost invariably equated to value by the assessor, the assessed value was 90% of price. The second sale was an REO disposition by the lender on the 1st deed of trust in spring 2009 for $350,000. At the time, prices in the neighborhood were approximately $400,000 for similar tract homes, all based on cash-equivalent financing. So, what's the drop in value: $250,000+ or $140,000+/-?

That leads to the second issue, condition. The buyers calculated that the sum of deferred maintenance on the place was on the order of $25,000. Using a perfectly good rule of thumb, "double your money", they adjusted their price accordingly. Now, in the greater scheme of things, $25000 of deferred maintenance on a foreclosure these days is NOTHING. I saw a place in Stockton with even the roof partially removed.

So, the premise that C-S compares apples to apples just simply isn't true. And, I hope the example shows how C-W probably overstates house value change 2005-9.