There is a very broad consensus the housing market has bottomed. The author of the article believes there is no convincing data to support the view that housing markets have turned the corner. He believes the housing recovery is actually a mirage and a new delinquency crisis is coming.
Apr. 8, 2013, 12:30 PM
Housing analyst Keith Jurow has been a regular contributor to Business Insider for almost three years. His new real estate subscription service – Capital Preservation Real Estate Report – will launch in a few weeks.
At the end of March, it was announced that the 20-city Case-Shiller Index was up 8.1% year-over-year. Nearly all housing experts declared that this was further confirmation that the housing recovery was firmly in place.
A few weeks earlier, Zillow had released its latest survey of 118 economists, strategists and other experts on the expected direction of home prices over the next five years. Every one expected higher prices with the most pessimistic prediction showing an average annual increase of 3%.
Clearly, there is a very broad consensus that the housing market has bottomed. In the article below, I review much of their analysis. For those of you who have never read any article of mine, I have been consistently asserting that there is no convincing data to support the view that housing markets have turned the corner.
Because my view is so out-of-sync with the consensus, it is worth taking an in-depth look at what my latest research has uncovered.
My conclusion is that the much-vaunted housing recovery is actually a mirage and that a new delinquency crisis is coming.
Home Prices in the Northeast
We’ll get to the Case-Shiller Index shortly. Let me begin by showing you raw sale price numbers from the largest family-owned brokerage firm in the northeast – William Raveis & Co. This table shows the average price-per-square-foot for single-family homes in five northeast states. The figures are for the months November 2012 through January 2013.
The figures show the average price-per-square foot (ppsf) for all single-family homes sold in these towns and cities. As far as I know, Raveis & Co. is the only brokerage firm in the country which provides these statistics. Their website also provides median sale prices. I wiIl explain why I prefer the use of average price-per-square-foot.
Here is why. A median price is simply the point at which half of the prices of homes sold are above it and half are below it. It is greatly influenced by the mix of homes sold. Sharp reductions in the sale of foreclosed properties will great affect the median price.
I’ve emphasized repeatedly in previous articles that the banks have severely curtailed the number of foreclosed properties (REOs) which they have put on the active MLS market. For example, in the spring of 2009, 2/3 of all homes sold in the Greater Phoenix area were foreclosed properties. In December 2012, a mere 10% of the homes sold were foreclosures.
Since foreclosed properties are normally the lowest-priced homes on the market, the sharp reduction in sales of these inexpensive homes will almost always raise the median price of sold homes. Yet a higher median price does not necessarily indicate that homes prices would be rising without this market manipulation by banks.
The average ppsf is not skewed by the plunge in foreclosed property sales nearly as much as the median sale price. But since foreclosed homes will normally have the lowest ppsf in any market, the average ppsf will also tend to be pushed up by the scarcity of foreclosures for sale. Thus it is all the more remarkable that prices in my home state of Connecticut showed double-digit declines in average ppsf for numerous towns.
Now you may object to the use of average ppsf. I have found that the ppsf does not vary all that much throughout a given town for comparable homes and does not change much month-to-month. Three bedroom homes tend to command a slightly higher ppsf than four bedroom houses. Nonetheless, I believe that average ppsf is a far more reliable indicator of home prices in a town than the median sale price.
Home prices in Connecticut have been much weaker than in the other four northeast states. Some of you will object that the figures from raveis.com show that prices are up in other New England cities and towns. Granted. But you can see that, with one glaring exception, year-over-year price increases in the towns I’ve listed have been tepid at best.
Cambridge, Massachusetts is the one bright spot. Remember that it is the home of both Harvard and MIT. As a college town, it has greater job stability than most other communities. So it is not at all representative of even the rest of Massachusetts.
If you think I may be cherry-picking towns, go to raveis.com and see for yourself. Just hit the drop-down menu called “Housing Data” and link to “View Local Housing Data.” You can view nearly every city or town in five states plus Westchester County in New York State.
My Concerns About The Case-Shiller Home Price Index
Okay, I can hear you say: What about the Case-Shiller 20-city Index, which has been showing increasing strength for many months?
That’s a very fair objection. Let’s examine the Case-Shiller Index – the gold-standard for home price indices. It clearly has more credibility than any other home price index in the nation. Standard & Poors – the publisher -- has actually put out a lengthy explanation of the methodology behind the Index which I have read more than once.
The Index uses what they call a “repeat sales” model because it takes recent home sales and matches them with a previous sale of the same property. You need to realize that certain important assumptions underlie the Index. By far, the most important is that very different weighting is assigned to matched pairs of home sales depending on certain questionable criteria.
A paired sale is assigned a weight which could be anywhere from zero to one depending on how far the pair differs from the “average price change for the entire market.” The purpose of this is to smooth out distortions which the creators believe are caused by extreme price changes that differ markedly from most of the other price changes in a given metro area.
The most dubious weighting factor is that a home which has a longer time interval between its two paired sales is given considerably less weight than one where the interval is much shorter. For example, a home in which the interval between sales is 15 years may be assigned a weight of only 70% of that of a paired sale with a nine-month interval. S & P explains that the weight could be as low as 55% of the paired sale with a 6-month interval.
Why does the Index weight home sales so differently? S & P explains the assumption behind the time interval weighting this way: “over longer time intervals, the price changes for individual homes are more likely caused by non-market factors” (i.e., physical changes in the property such as adding a bedroom).
I am convinced that this weighting of sales is a real flaw in the accuracy of the Index. In its 2012 Profile of Home Buyers and Sellers, the National Association of Realtors reported that the median length of time that home sellers had owned their house was 9 years. This means that nearly every home sold in 2011-2012 was given a much lower weighting than the few homes which were resold within two years or less.
Why is this so important? The different weighting given to paired sales causes the index figure to be very far removed from the raw sales data. You may think that this weighting is justified. I prefer sales data to be as close to the raw sales price as possible to avoid distortions. As I see it, the weighting used to create the Case-Shiller Index distorts the price changes in the major housing markets in the Index. If its methodology were used on the raveis.com numbers for New England, their sale prices would look completely different.
To see entire article CLICK HERE