Monday, June 25, 2012

The Housing Recovery is one year old outside the Case Shiller cities
Measured by the FHFA (Federal Housing Finance Administration) Index, housing is flatting out, but prices in Case Shiller cities are still falling.  The Case Shiller 20 city index is reasonably accurate, but unrepresentative.  House prices in these cities are twice the national level.  While the Case Shiller index declined 34% since the peak in 2006, prices outside these regions declined only 14%.  FHA data for the 20 cities in the Case Shiller is close to the Case Shiller index. Both indexes are accurate they just measure different geography.  
This is summarized below

Heartland prices bottomed out in the first quarter of 2011 and they have been increasing through the first quarter of 2012. In the last three months, Case Shiller cities benefited from reinstatement of higher FHA loan limits which is a one-time pop. 
So in the heartland prices are improving but the Case Shiller cites in California and the Northeast (New York and Boston) are falling.  The US housing market in recovery will differ sharply by region which it always does.  The recovery is good news unless you own property in cities like Los Angeles which is 213% above the start of the bubble and has a substantial decline ahead.  Dallas, a Case Shiller city where the prices are 128% above start of the bubble, is similar to a heartland city.  Los Angles fell 40% since the peak while Dallas fell 5%.  Adjusted for inflation Heartland regions have lost 18% off their value the Case Shiller cities gained 20%.  The differences are unsustainable.

Thursday, June 21, 2012

Case Shiller results are a onetime pop due to changed FHA loan limits

The May Case Shiller 20 city index comparison to prior year shows encouraging signs of recovery.  Case Shiller recommends this comparison contending that the seasonally adjusted index is misleading.  It is.  Working with seasonally adjusted data brings up unexpected problems and unpublished adjustments, but as this example demonstrates, without understanding monthly data, it is impossible to understand price trends.
The following table illustrates the effect of changes in FHA loan limits.  Los Angeles prices in fourth quarter 2011 fell at an annual rate of 10%, but  first quarter 2012 prices increased by 1%, an improvement of 11 points between quarters.  Los Angeles prices fell 40% from June 2006 to March 2012 but they still averaged $519,000.  Price trends in 18 of the 20 cities improved over the prior quarter while 2 low priced cities remained stable.  This sounds too good to be true, and it is.
May data covers house closings in March through January 2012.  The purchase commitment is usually made two months earlier from November to January.  Thus index publication lags purchase commitment by an average of 5 months – enough time to forget what happened 5 months earlier
The FHA loan limits which were raised to a maximum of $729,750 to ease the housing crisis were cut to the normal level of $625,000 on October 1, 2011.  As October approached buyers and sellers rushed to close while FHA financing was available, further lowering prices.  Commitments made from May to July declined at an annual rate of 5%.  The decline accelerated to 8% in the next three months.  Lobbying by real estate interests succeeded in reinstating the $729,750 limit plus increasing the limit in the lower priced regions by 10 points to 125% of the median price on November 18, 2011.  As a result the next three months, the 20 city index was flat - an annual improvement of 9 percentage points.
The highest priced markets received the greatest increase.  In volatile markets with average prices of $519,000 the quarterly price change improved at an annual rate of 10%.  In the lower priced stabilized markets with an average price of $184,000 the improvement was only 3%.  Higher loan limits have the greatest impact in high priced markets.  Markets with the largest decline have the greatest improvement.  Excluding, Detroit which was facing bankruptcy and Atlanta where the high unemployment is leading to high foreclosures, the annual rate improvement has a 75% correlation to the fall in prices.  Of course, this annual rate only occurred for one quarter so the 9 percent annual rate is 2.9% per quarter.  
The first quarter results will change when data for two months of March and one month of February is added to the index.  Usually these adjustments are small, but changes in loan limits have resulted in substantial adjustments.  The Case Shiller cities will resume declining as they were before the loan limit change.  With 3.2% down payment for FHA mortgages where the FHA loses 64% of value on foreclosure, it takes only a small decline to make this a big loser for the taxpayers.  More generous loan limits provide a short lived improvement.  Although it is difficult to forecast the past, when May price negation data is available at Halloween, the index will be scary.
The good news is that housing markets outside the Case Shiller 20 cities are recovering. These markets are less than half the price of these and much more stable.