Wednesday, June 12, 2013

Case Shiller obscures housing reality

“We require 4 acre lots. We won’t extend the sewers so a septic system  must be built. One will do, but we require 2, because requiring 2 systems increases the chance that the lot will fail the perk test (for the septic) so the permit can be rejected.  The building codes are excessive. Permitting can drag on for years. But people are still ticked off at me for letting new houses get built. Legally you can’t stop someone from building on land they own.  All I can do is make it as expensive and drawn out as possible.”  A locality that did this would see prices soar because new homes could not be built to meet demand.  New houses are all huge and to not fit with the smaller existing homes in the town.  This comment is from the mayor of a town in the New York Metropolitan Statistical Area. (All Case Shiller data is for the MSA.)
This policy is unusual, but such localities drive the Case Shiller index.  This index is weighted by housing wealth. New York, Los Angeles, San Diego, and San Francisco alone make up 47% of the index. Different locations use different methods of controlling growth. In California, which added the Coastal Commission to its land use controls, prices at the 2006 peak reached 337% of prices at the start of the bubble- well above the Case Shiller average of 258%.  The areas outside the Case Shiller 20 which I call Heartland, contains 80% of the national housing units. These markets were affected by the euphoria about the gains in the hot markets and reached 150% of the price at the start of the bubble.

The Federal Housing Finance Administration (FHFA) index is used as the national index.  This is a matching index with a two month lag like Case Shiller but the calculation details differ. The index formerly contained only homes backed by Fannie Mae and Freddy Mac.  The Current Expanded index adds data for FHA and for county property recorders to cover home that are too expensive for the government programs. The FHFA data was used to recalculate 20 Case Shiller cities index, which was weighted and subtracted from the FHFA national data to form an index of the regions outside the 20 Cities - the Heartland Index. The expanded data is available for the total and major MSA and MSA divisions, while the smaller MSA’s only have Fannie May and Freddy Max data.  (Data on large cities is reported by several divisions some of which have only Fannie and Freddy data.) The 20 cities index using FHFA data from January 2012 ranged from 4.4 %t to % to 5.9% below the Case Shiller data.  Not precise, but the Heartland   index is the most accurate indicator of the market outside the Case Shiller.
Case Shiller produces a national number which moves closely with the 20 City Index, A comparison of the National Index with the Case as shown on the first page of their monthly press release. This indicates that the Case Shiller 20 is very close the national trends.   If you read the methodology and understand housing data you can see this is absurd.  It does not match the National FHFA numbers because it does not cover the lower priced 37% of the country.  For example it covers 72% of New York which is the New York City MSA in the State.  In Texas they cover 54% which is the Dallas MSA. California has 98% coverage. They only cover large Cities and some other odd expensive markets. It is a biased sample that misrepresents national housing. 
 The Following Chart illustrates the Housing indexes adjusted for inflation

At end of a contraction housing tends to return to the inflation adjusted average price at the beginning of the bubble. The Case Shiller is 27% above that level while the Hartland  Index is  8% below.
Federal programs first impacted prices in the second quarter of 2009.  These actions such as the house purchase credit, higher FHA loans limits, mortgage assistance and lower interest rates increased the Case Shiller 20, but had little impact on heartland markets. Since the programs the gap between the case Shiller 20 and the rest of the country has widened. Within the Case Shiller, those cities that had the highest increases in the bubble had the highest increase.
Demand is high in the overheated markets but the new homes are very expensive and inadequate to meet demand that was created by low mortgage rates and FHA loans.  It is easy to build homes in the Heartland and 75% of permits are issued in these regions but with 10 months of inventory demand is relatively weak.  The bubble has returned to California and a few other markets where investors believe the market will return to the peek. That will persist as long as interest rates are low but in the long term it is unsustainable.  The Heartland is recovering and it is very sustainable.
The Case Shiller is widely believed.  That is why the media talks about the shortage of houses for sale when the March Inventory is at 9* months of sales nationally compared to 3 months in California. The belief that Case Shiller index reflects what is going on in the country illustrates just how poorly housing is understood. 
*Source Zillow data